TMS for OCD: What Is It? How Transcranial Magnetic Stimulation Works and What the Evidence Shows
Transcranial magnetic stimulation, or TMS, is a noninvasive neuromodulation treatment that uses rapidly changing magnetic fields to induce small electrical currents in targeted cortical tissue. In obsessive-compulsive disorder (OCD), repeated stimulation is intended to change activity within brain networks involved in obsessive thoughts, compulsive behavior, cognitive control, error signaling, and habit-like responding. TMS does not diagnose OCD, erase thoughts, or directly prevent a person from performing compulsions. It is a treatment intervention aimed at neural circuits that contribute to the disorder.
The short answer to the question “does TMS work for OCD?” is that active TMS has outperformed sham stimulation across randomized trials and meta-analyses, but the average benefit is modest, protocols are heterogeneous, and response is not guaranteed. The strongest regulatory evidence in the United States supports specific repetitive and deep-TMS protocols rather than every machine, coil, target, frequency, or clinic-defined version of “TMS for OCD.” The 2026 OCD clinical-practice guideline describes rTMS as an augmentation option for resistant OCD while emphasizing that the evidence continues to evolve.
That distinction matters because TMS sits in a broader treatment pathway. Evidence-based OCD treatment still centers on specialized cognitive behavioral therapy, particularly exposure and response prevention (ERP), and serotonin-reuptake-inhibiting medication. TMS is generally considered when symptoms remain clinically significant despite established treatment, when an additional noninvasive option is appropriate, or within specialist care. It should be understood as one component of treatment planning rather than as a replacement for careful diagnosis, ERP expertise, medication review, and measurement of actual functional improvement.
This article explains what TMS is, how repetitive TMS and deep TMS differ, what the FDA has cleared for OCD, what randomized trials and meta-analyses actually show, how symptom provocation fits into treatment, what a course typically involves, who may be considered, the main safety issues, and how TMS compares with ERP, medication, electroconvulsive therapy, and deep brain stimulation.
What is TMS for OCD?
TMS delivers magnetic pulses through a coil placed on or above the scalp. A pulse creates a changing magnetic field that passes through the skull and induces an electric field in nearby brain tissue. When pulses are delivered repeatedly, the procedure is called repetitive transcranial magnetic stimulation, or rTMS. Repeated sessions can influence cortical excitability and network function beyond the brief moment of stimulation.
For OCD, investigators have studied several cortical targets because OCD involves distributed cortico-striato-thalamo-cortical networks rather than a single “OCD center.” Common research and clinical targets include the medial prefrontal cortex and anterior cingulate cortex (mPFC/ACC), dorsolateral prefrontal cortex (DLPFC), supplementary motor area and pre-supplementary motor area (SMA/pre-SMA), and orbitofrontal cortex. Different targets can also be paired with different frequencies and coil designs, so two treatments both called “TMS for OCD” may be physiologically and clinically different interventions.
The term deep TMS, often shortened to dTMS, refers to systems using coil designs intended to stimulate a broader and relatively deeper cortical volume than many conventional focal coils. “Deep” does not mean that electrodes are implanted into the brain. Deep TMS remains noninvasive. This separates it fundamentally from deep brain stimulation (DBS), which is neurosurgery involving implanted electrodes and is reserved for a much narrower group of people with severe, highly treatment-refractory OCD.
Is TMS FDA-cleared for OCD?
Yes, specific TMS systems and protocols have received U.S. Food and Drug Administration clearance for OCD. In August 2018, the FDA permitted marketing of the BrainsWay Deep TMS System for OCD after reviewing a multicenter randomized sham-controlled study. The FDA's current device database continues to list BrainsWay OCD systems under product code QCI, the Class II classification for transcranial magnetic stimulation systems for obsessive-compulsive disorder. Current FDA device listing is the appropriate place to verify a marketed system's classification and submission history.
Other systems have subsequently received OCD indications through the FDA's 510(k) pathway. For example, the MagVenture submission K193006 was cleared in 2020 for adjunctive treatment of adult patients with OCD. The practical lesson is precise: FDA clearance applies to particular devices and labeled protocols. It is not a blanket statement that every form of rTMS, every cortical target, every frequency, or every off-label protocol has equivalent evidence.
FDA clearance also answers a different question from a clinical guideline. A device can meet U.S. regulatory requirements for a specified indication while guideline organizations weigh the entire literature differently when deciding how strongly to recommend a procedure in routine care. That helps explain why U.S. availability and international recommendations do not always look identical.
How does TMS work in OCD?
The most defensible mechanism-level description is network modulation. OCD is associated with altered function and connectivity across cortico-striato-thalamo-cortical circuits that include frontal and medial cortical regions, the anterior cingulate cortex, striatum, and thalamus. TMS acts at the cortex but can influence connected networks. The goal is therefore not to “turn off” a single brain area. It is to alter the dynamics of a circuit whose activity is associated with persistent obsessions, compulsive responding, threat and error processing, and difficulty disengaging from repetitive behavioral loops.
Frequency is relevant but cannot be interpreted in isolation. Lower-frequency stimulation is often described as relatively inhibitory and higher-frequency stimulation as relatively excitatory, yet the clinical effect depends on target, baseline brain state, coil geometry, intensity, pulse pattern, session number, and network connectivity. OCD trials have reported benefit with both higher- and lower-frequency approaches at different targets. The 2024 network meta-analysis by Vinod and colleagues found evidence favoring several distinct strategies, including excitatory bilateral DLPFC stimulation, inhibitory right DLPFC stimulation, excitatory or inhibitory bilateral mPFC/ACC stimulation, and inhibitory bilateral SMA stimulation.
This heterogeneity is one reason a clinic's statement that it “offers TMS” is not enough information. For OCD, a patient should know which device, coil, anatomical target, frequency, intensity, pulse count, course length, and symptom-provocation procedure are being used, and whether the protocol is FDA-cleared for OCD, evidence-supported but off-label, or experimental.
How is deep TMS different from conventional rTMS?
rTMS is the broader category: any TMS protocol that delivers repeated magnetic pulses. Deep TMS is one implementation within that broader family. Conventional figure-of-eight coils are comparatively focal, while H-coils and other designs used for deep TMS distribute the induced field differently and can engage a broader region. For the pivotal FDA-cleared OCD protocol, stimulation was designed to engage the dorsal medial prefrontal cortex and anterior cingulate network bilaterally.
The distinction should not be converted into a simple hierarchy in which “deeper” automatically means “better.” Depth, focality, field distribution, target, and protocol all matter. Meta-analyses of OCD trials find positive signals across several conventional and deep-TMS approaches, but direct evidence does not establish one universal protocol as best for every person. Network meta-analysis can rank protocols statistically, yet rankings are limited when many underlying trials are small and few studies test exactly the same intervention.
Where does TMS fit in the OCD treatment pathway?
For most people, TMS is not the starting point. The first treatment question is whether the person has received a valid OCD diagnosis and differential assessment and an adequate trial of established treatment. OCD-specific CBT with ERP has a strong evidence base. Medication treatment commonly uses serotonin reuptake inhibitors, with clomipramine and other strategies considered according to clinical context. When a partial medication response persists, specialist clinicians may consider strategies such as antipsychotic augmentation for selected patients.
The 2026 clinical-practice guideline update keeps SSRIs and CBT as first-line treatments and places rTMS in the augmentation pathway for resistant OCD. The same guideline notes substantial methodological heterogeneity in neuromodulation studies. This positioning is useful because it prevents a common category error: evidence that TMS can reduce OCD severity does not imply that TMS has replaced ERP or medication as the default first intervention.
The phrase treatment-resistant OCD also needs clinical context. It does not mean “I tried one thing and still have symptoms,” and it is not a separate diagnosis. Resistance is judged from the adequacy, duration, dose or intensity, adherence, and quality of prior treatments, alongside the certainty of the diagnosis and the presence of comorbid conditions. A specialist may also identify pseudo-resistance, such as medication that was never used at an adequate dose or duration, ERP that did not actually include response prevention, treatment disrupted by severe avoidance or family accommodation, or symptoms attributed to OCD that are better explained by another condition.
What does a course of TMS for OCD involve?
The exact course depends on the system and protocol, but treatment generally begins with a psychiatric and safety assessment, confirmation of the clinical target and indication, review of medications and seizure-risk factors, and measurement of baseline OCD severity. The Yale-Brown Obsessive Compulsive Scale (Y-BOCS) is frequently used to quantify severity and change in clinical trials and clinics. A Y-BOCS score is an outcome measure, not a stand-alone diagnostic test; diagnosis depends on a clinical assessment of obsessions, compulsions, distress, impairment, duration, differential diagnoses, and context.
Motor-threshold measurement
TMS intensity is typically calibrated relative to an individual's motor threshold: the minimum stimulation intensity that reliably produces a motor response under standardized conditions. In the pivotal deep-TMS OCD protocol, intensity was based on the leg motor threshold because of the coil position and intended medial-frontal target. Motor-threshold procedures allow dose to be individualized to the person's neurophysiologic response rather than using one absolute machine output for everyone.
Targeting and coil placement
The coil is positioned according to the selected protocol. In the pivotal H7-coil study, the target was the dorsal medial prefrontal cortex/anterior cingulate region, and the coil was positioned anterior to the motor location used for calibration. Other research protocols target DLPFC, SMA/pre-SMA, or other frontal areas. Some centers use neuronavigation; others use protocol-defined scalp measurements or device-specific targeting methods.
Symptom provocation before stimulation
A distinctive feature of the FDA-cleared OCD approach is brief individualized symptom provocation before and during stimulation. The clinician identifies triggers likely to activate the person's OCD network, such as a contamination cue, uncertainty statement, responsibility scenario, forbidden-thought cue, incompleteness trigger, or other obsession-relevant prompt. The goal is to evoke a meaningful but tolerable symptomatic state immediately before stimulation.
Symptom provocation is related to behavioral principles but should not be casually equated with a complete ERP session. ERP is a structured psychological treatment in which a person intentionally confronts obsession-triggering situations while reducing ritualized responses and learning new ways of relating to uncertainty, distress, and urges. The provocation used in TMS is brief and is designed primarily to place the targeted circuit into a symptom-relevant state during stimulation.
The rationale is state-dependent neuromodulation: the effect of stimulation may depend partly on which network is active at the time. A 2025 JAMA Psychiatry systematic review and meta-analysis found that active TMS outperformed sham in OCD studies both with and without symptom provocation. The estimated additional effect attributable to provocation was not statistically significant, which means the current literature does not prove that provocation itself causes the added clinical benefit. It remains part of cleared OCD protocols and a plausible mechanistic component, while the causal evidence is still developing.
Stimulation session and schedule
The pivotal deep-TMS study used high-frequency 20 Hz stimulation at 100% of the leg resting motor threshold, delivered in 2-second trains separated by 20-second inter-train intervals, with 50 trains and 2,000 pulses per session. Sessions were delivered daily on weekdays over a six-week acute course. Device labeling and later cleared systems can specify their own exact parameters, so this protocol should be treated as a concrete example rather than as instructions for self-directed treatment.
A person remains awake during treatment. TMS does not normally require general anesthesia, an operating room, or surgical implantation. The magnetic pulses create loud clicking and can produce tapping, pressure, scalp or facial muscle contractions, or discomfort. Hearing protection is used. After a routine session, many people can resume ordinary activities, subject to the treating clinician's instructions and any individual reaction.
What does the best randomized trial show?
The pivotal multicenter trial by Carmi and colleagues (2019) is central to the evidence base for FDA-cleared deep TMS in OCD. Ninety-nine adults with OCD were treated at 11 centers and randomized to active or sham high-frequency deep TMS targeting the medial prefrontal/anterior cingulate region. Treatment was delivered after individualized symptom provocation for six weeks.
At the end of treatment, the mean reduction in Y-BOCS score was 6.0 points with active deep TMS compared with 3.3 points with sham. Response was defined as at least a 30% reduction in Y-BOCS score. By that threshold, 38.1% of the active-treatment group responded compared with 11.1% of the sham group. At the one-month follow-up, response rates were 45.2% and 17.8%, respectively.
Those figures support a real treatment signal, but they should not be marketed as a universal “success rate.” They come from one protocol, a defined trial population, a particular response threshold, and controlled study conditions. They also show an important clinical reality: most participants did not meet the prespecified response threshold immediately after the acute course, even though the active group did substantially better than sham.
What do systematic reviews and meta-analyses show?
The broader evidence base supports efficacy while also exposing uncertainty about magnitude and optimal protocol. A 2022 systematic review and pairwise/network meta-analysis included 21 sham-controlled studies with 662 participants. Across protocols, rTMS favored active treatment with a pooled Hedges' g of -0.502. In protocol-specific network analyses, low-frequency pre-SMA stimulation, high-frequency bilateral DLPFC stimulation, and low-frequency right DLPFC stimulation each showed significant benefit. The authors emphasized that the evidence network was sparse, most trials were small, protocols were heterogeneous, and publication bias was a concern.
A 2023 systematic review and meta-analysis by Grassi and colleagues pooled 31 trials and reported a significant overall effect of active rTMS on OCD symptoms, with subgroup signals across bilateral pre-SMA, DLPFC, mPFC/ACC, and orbitofrontal targets. Again, the main implication is not that every target works equally. It is that therapeutic effects have been observed across several network nodes while the field continues to refine which protocol is most reliable for which patient.
A 2024 network meta-analysis likewise found several protocols superior to sham and concluded that medial and lateral prefrontal targets were promising for resistant OCD. The authors also noted modest sample sizes and risk-of-bias concerns in much of the literature. Network ranking is useful for hypothesis generation and comparative inference, but it cannot substitute for large head-to-head trials.
For deep TMS specifically, a 2024 meta-analysis by Li and colleagues included four randomized controlled trials with 252 participants described as having treatment-resistant OCD. Active deep TMS produced a higher Y-BOCS response rate than sham immediately after treatment, with a risk ratio of 3.71, and at one-month follow-up, with a risk ratio of 2.60. No serious adverse events were documented in the included studies. The same paper explicitly noted that the high-quality evidence base remained small, an important limitation when interpreting large relative effects.
The most recent broad synthesis in this evidence set, a 2025 systematic review, meta-analysis, and meta-regression by Figueiredo and colleagues included 31 trials. Across repeated rTMS protocols, the pooled mean difference in Y-BOCS score was -3.30 points versus sham, with a 95% confidence interval from -4.55 to -2.04. The authors compared this with a reported minimal clinically important difference of 4.9 points, meaning the average pooled advantage was statistically significant but smaller than that clinical benchmark. The estimated number needed to treat for response was 6, and dropout rates were comparable with sham. Subgroup signals were larger for left DLPFC and mPFC/ACC stimulation.
These results can coexist without contradiction. A trial can show a higher probability of meeting a categorical response threshold while the average between-group change remains modest. Relative response ratios can look large when sham response is low. Network meta-analyses can identify promising targets even when direct comparative data are sparse. And a person can experience a clinically meaningful response even if the average effect across all participants falls below a chosen minimal-important-difference threshold.
What is the realistic success rate of TMS for OCD?
There is no single scientifically valid percentage that applies to every patient and protocol. The most defensible figures should be tied to their study design. In the pivotal randomized deep-TMS trial, 38.1% met the study's response definition immediately after treatment and 45.2% did so at one month, compared with 11.1% and 17.8% under sham. That is randomized evidence for a specific high-frequency mPFC/ACC deep-TMS protocol.
Real-world observational data have reported higher rates. In a 22-site post-marketing study by Roth and colleagues, 219 patients contributed treatment information and 167 with at least one post-baseline Y-BOCS measurement were included in the main analyses. The investigators reported 72.6% reaching a first response at some point, 52.4% reaching a sustained response for at least one month, and a 57.9% response rate among patients with Y-BOCS scores after 29 sessions.
Those observational numbers should not be treated as a replication of the randomized trial's effect size. There was no randomized sham comparison, outcome availability differed across patients, clinical practice conditions varied, and several authors had manufacturer affiliations. Real-world evidence is useful for understanding what happens in practice, but it is more vulnerable to selection, measurement, expectancy, concomitant-treatment, and follow-up biases than a blinded randomized trial.
“Response” is also not the same as remission, cure, or complete recovery. Many TMS studies define response as a percentage reduction in Y-BOCS score, often 30%. A person can cross that threshold and still have clinically significant symptoms. Conversely, a person can improve meaningfully in functioning without crossing a particular research cutoff. Good clinical monitoring looks at severity, time consumed, avoidance, distress, compulsions, work or school function, relationships, and quality of life rather than relying on one percentage alone.
Why do TMS studies use different targets?
The diversity of targets reflects both the biology of OCD and the history of TMS research. OCD involves multiple interacting cortical and subcortical loops. Researchers therefore test stimulation at accessible cortical nodes that are connected to broader OCD networks. The mPFC/ACC approach emphasizes medial frontal circuitry related to error, conflict, affective salience, and action selection. DLPFC protocols emphasize cognitive-control and prefrontal network regulation. SMA/pre-SMA protocols target regions implicated in motor preparation, action inhibition, and repetitive responding.
A positive result at more than one target does not mean target selection is irrelevant. Different protocols can produce different electric-field distributions and different downstream network effects. The present literature supports several plausible targets but does not yet provide a validated biomarker that tells clinicians exactly which target will work best for a particular individual. Personalized connectivity-guided approaches are an active research direction rather than a settled standard for OCD care.
Does high-frequency or low-frequency TMS work better?
No single frequency can currently be declared universally superior for OCD. The successful FDA-cleared deep-TMS protocol uses high-frequency 20 Hz stimulation of the mPFC/ACC network. At the same time, meta-analytic evidence supports low-frequency stimulation at right DLPFC and pre-SMA/SMA targets, while some analyses also find benefit from high-frequency bilateral DLPFC stimulation. Frequency and target must therefore be interpreted together.
Theta-burst stimulation is also being studied because it can deliver patterned stimulation in shorter sessions. Evidence in OCD remains less established. In the 2025 Figueiredo meta-analysis, continuous theta-burst stimulation did not show the same efficacy signal seen with several repeated-rTMS approaches. That result does not close the research question, but it is a reason not to assume that a faster or newer pulse pattern is automatically equivalent to the best-supported OCD protocols.
Does symptom provocation make TMS more effective?
The answer is plausible but not proven. Symptom provocation is built into FDA-cleared OCD protocols because TMS effects can depend on the state of the stimulated network. Activating an obsession-related state before stimulation may make the relevant network more available to modulation. This is conceptually coherent and clinically operationalized.
However, the Bello et al. 2025 meta-analysis provides the most useful quantitative caution. For OCD, active TMS beat sham both in studies using provocation and in studies without it. The estimated additional effect associated with provocation was an SMD of -0.22, with a 95% confidence interval from -0.65 to 0.20, and was not statistically significant. The authors concluded that direct comparisons of TMS with and without provocation are needed to establish a causal effect.
Practically, this means a clinic using an FDA-cleared provocation-based protocol should implement the provocation competently rather than omit it casually. It also means patients should not be told that the provocation component itself has independently proven superiority. The current evidence supports TMS efficacy more strongly than it establishes the incremental contribution of provocation.
What are the side effects of TMS for OCD?
The most common adverse effects are local and transient: headache, scalp discomfort, application-site pain, facial or jaw discomfort, muscle twitching, and neck discomfort. In the FDA's 2018 review headache was reported by 37.5% of patients receiving active BrainsWay treatment and 35.3% receiving sham, while other reported local adverse effects were generally mild or moderate and resolved shortly after treatment. The pivotal trial did not identify a serious device-related adverse-reaction signal.
TMS produces loud clicks, so hearing protection is part of standard treatment. Scalp and facial sensations can be strongest early in a course and may become easier to tolerate as the person becomes familiar with the procedure or as clinicians make protocol-permitted adjustments.
The most important rare neurologic risk is a seizure. Modern TMS has a low seizure risk when delivered within established safety parameters, but risk is not zero. The international expert safety guidelines by Rossi and colleagues emphasize pre-treatment screening, attention to stimulation parameters, medications and substances that can alter seizure threshold, sleep deprivation, neurologic history, and device-specific precautions. A history of seizures does not automatically answer candidacy by itself; it requires individualized medical assessment.
Mood activation is another consideration. TMS can affect mood networks, and clinicians should screen for bipolar-spectrum history and monitor for clinically meaningful mood elevation, agitation, or behavioral change. The probability and relevance depend on the individual, co-occurring conditions, medications, and protocol. New severe symptoms during a course should be evaluated rather than assumed to be an expected part of treatment.
Who should not receive TMS, or needs additional safety review?
Safety screening focuses especially on metal and electronic implants in or near the head because strong, rapidly changing magnetic fields can interact with certain materials or devices. The FDA's original OCD authorization lists contraindications for the BrainsWay system that include certain metallic objects and implanted stimulator devices in or near the head, such as cochlear implants, deep-brain stimulators, vagus-nerve stimulators, implanted electrodes, aneurysm clips or coils, stents, and metallic fragments. Exact contraindications are device-specific, so a clinic should use the current manufacturer's labeling rather than a generic internet checklist.
Additional review is appropriate for seizure history, significant neurologic disease, medications or substances that may alter seizure threshold, pregnancy, unstable medical illness, and any implanted medical device. Dental fillings and ordinary orthodontic or surgical materials are not all equivalent risks; compatibility depends on location, composition, fixation, and the TMS system. Patients should give the treatment team a complete implant and medical history instead of trying to decide compatibility on their own.
Who may be a candidate for TMS for OCD?
A reasonable evaluation begins with a confirmed OCD diagnosis, clinically significant current symptoms, and a review of treatment history. For many patients, TMS enters the conversation after an adequate course of specialist ERP and adequate medication trials have not produced sufficient improvement, have produced only partial improvement, or cannot be used as planned because of clinically important tolerability or medical constraints. Insurance or local regulatory criteria may be narrower than a clinician's scientific judgment.
Candidacy is not determined by symptom theme. Contamination OCD, checking, harm-related obsessions, symmetry or incompleteness symptoms, sexual or religious obsessions, relationship themes, and other presentations can all occur within OCD. What matters more is whether the diagnosis is correct, symptoms are sufficiently severe or impairing to justify the intervention, the person has an appropriate treatment history, the selected protocol fits the indication, and there are no unacceptable safety problems.
Comorbidity also requires interpretation rather than automatic exclusion. Depression, anxiety disorders, tic disorders, ADHD, autism, trauma-related conditions, substance use, or bipolar-spectrum illness can affect treatment priorities, measurement, and risk. If the presenting problem is not actually OCD, an OCD TMS protocol may target the wrong clinical problem. This is another reason the diagnostic assessment must precede device selection.
Can TMS be used with ERP?
Yes. TMS and ERP act through different but potentially complementary treatment processes. ERP directly changes the person's learning and behavior around obsessional triggers, uncertainty, rituals, avoidance, and safety behaviors. TMS attempts to modulate neural circuits implicated in OCD. In practice, many patients continue psychotherapy during a TMS course, although scheduling and treatment sequencing vary.
A key point is that the brief symptom provocation used in some TMS protocols is not a substitute for a full course of ERP. Someone can receive technically correct TMS while still needing systematic work on compulsions, reassurance seeking, avoidance, mental rituals, family accommodation, and relapse-prevention skills. If TMS lowers symptom intensity enough to make ERP more feasible, that may be clinically useful even when TMS is not sufficient as a stand-alone intervention.
Can TMS be used with OCD medication?
Often, yes. The pivotal randomized deep-TMS trial allowed participants already receiving OCD medications to remain on stable regimens, which means the evidence is compatible with an adjunctive-use model rather than requiring medication discontinuation. The OCD medication overview explains the broader role of SSRIs, clomipramine, benefits, adverse effects, and monitoring.
Medication changes during a TMS course can complicate interpretation of response and can alter safety factors such as seizure threshold. A treating team may therefore prefer medication stability when clinically appropriate. That is not a universal rule to stop, start, or freeze medication. Medication management remains individualized, especially when a person has severe symptoms, comorbid depression or bipolar-spectrum illness, adverse effects, or other medical considerations.
TMS vs ERP for OCD
ERP has the more established role as a first-line psychological treatment for OCD and directly targets the behavioral reinforcement cycle that keeps compulsions going. TMS is a neuromodulation intervention generally positioned later or as an adjunct. They are therefore not interchangeable versions of the same treatment.
If someone has never received competent OCD-specific ERP, the clinically important question is usually why not. Access barriers, fear of exposure, previous generic therapy mislabeled as ERP, inadequate response prevention, severe comorbidity, or inability to tolerate treatment may all matter. TMS can be considered within this context, but the existence of TMS does not remove the need to evaluate whether first-line therapy was actually delivered adequately.
TMS vs medication for OCD
Medication and TMS differ in route, burden, adverse-effect profile, and evidence history. SSRIs and clomipramine have decades of evidence and guideline use in OCD. Medication exposes the whole body to a pharmacologic agent and can produce systemic adverse effects; TMS is a localized noninvasive procedure but requires repeated clinic visits and has protocol-specific neurologic and device-safety considerations.
A person with persistent symptoms may receive TMS while continuing medication rather than choosing one or the other. The treatment decision should be based on prior response, dose and duration of medication trials, adverse effects, ERP history, symptom severity, medical factors, patient preference, access, and the quality of the available TMS protocol. A clinic should not label TMS “drug-free” in a way that implies evidence-based medication must be abandoned.
TMS vs ECT for OCD
TMS and electroconvulsive therapy (ECT) are fundamentally different procedures. TMS uses magnetic pulses to stimulate cortical networks without general anesthesia and without intentionally inducing a generalized seizure. ECT is performed under anesthesia and intentionally induces a therapeutic seizure. ECT has a major evidence-based role in severe mood disorders and some other psychiatric emergencies, but it is not an established treatment for core OCD symptoms.
The 2026 OCD guideline states that ECT has no proven value for treatment-resistant OCD itself, while noting that it may be relevant when a severe comorbid condition such as depression has an independent ECT indication. That distinction prevents improvement in comorbid depression from being misrepresented as evidence that ECT directly treats OCD.
TMS vs deep brain stimulation for OCD
TMS is noninvasive and delivered from outside the skull. DBS requires neurosurgical implantation of electrodes connected to a pulse generator. Their risk profiles, candidacy thresholds, reversibility, cost, evidence base, and clinical infrastructure are therefore very different.
DBS is considered only for a small subset of people with severe, chronic, disabling OCD that remains refractory after extensive evidence-based treatment. TMS can be considered much earlier because it does not require brain surgery. A person comparing the two should read the dedicated deep brain stimulation for OCD guide rather than treating DBS as simply a stronger version of TMS.
How long does it take for TMS to work?
Improvement is usually evaluated across a multiweek course rather than after one or two sessions. In the Roth real-world dataset, the average first response occurred after about 18.5 sessions among those who reached that response definition, and sustained response began at about 20 sessions on average. The pivotal randomized trial assessed its primary outcome after a six-week course, with additional response observed at one-month follow-up.
These averages should not be turned into a deadline. Some people improve earlier, some later, and some do not achieve a clinically meaningful response. Repeated symptom measurement helps distinguish genuine change from day-to-day fluctuation. A clinic should define in advance how it will measure response, when it will reassess the treatment plan, and what counts as sufficient improvement to continue, extend, modify, or stop the course.
How durable are the benefits?
Durability is one of the less settled parts of the evidence base. The pivotal deep-TMS trial showed that the group difference persisted and the proportion meeting response criteria increased at one month. The 2024 deep-TMS meta-analysis also found a response advantage at one-month follow-up. The 2025 Figueiredo meta-analysis found no significant deterioration between end-of-treatment and the follow-up points available in the included trials.
Longer-term evidence is thinner, and there is no single standardized maintenance-TMS schedule for OCD supported across devices and populations. Some clinics offer tapering, booster, or maintenance sessions, but frequency and duration vary. Any maintenance plan should be presented as protocol- and evidence-specific rather than as an automatic requirement or a guaranteed way to prevent relapse.
Can TMS make OCD worse?
Transient distress can increase during symptom provocation because provocation is designed to activate obsession-related discomfort. Local physical side effects can also make sessions unpleasant. Those experiences are not the same as a sustained worsening of the disorder. Clinicians should still monitor OCD severity, functioning, mood, sleep, agitation, suicidality, and new neurologic symptoms throughout treatment rather than assuming every change is harmless.
A meaningful sustained worsening, emergence of mania or severe agitation, seizure-like event, new neurologic symptom, or major change in suicidality requires prompt clinical assessment. OCD itself can include intrusive thoughts about harm or suicide that are ego-dystonic obsessions, while suicidal intent is a different clinical construct; treatment teams must assess that distinction directly when it is relevant instead of inferring risk from the topic of an intrusive thought alone.
What happens if TMS does not work?
Nonresponse should trigger reassessment, not a reflexive escalation to the most invasive option. The team should confirm the diagnosis, review whether the selected TMS protocol was appropriate and adequately delivered, examine adherence and session completion, reassess ERP quality and medication adequacy, and identify comorbid conditions that may be driving impairment. Measurement matters because “I feel no different” and “the Y-BOCS fell but daily functioning did not improve” point to different clinical problems.
Depending on the case, next steps may include higher-quality or more intensive ERP, medication optimization, switching or augmenting medication, structured intensive treatment, another evidence-supported neuromodulation protocol, or specialist evaluation for advanced interventions. For medication-resistant partial response, antipsychotic augmentation is one evidence-based pharmacologic topic. For the rare patient with profound, chronic, highly refractory illness, specialist programs may eventually consider invasive neuromodulation such as DBS.
Why do FDA clearance and NICE guidance look different?
The United States and the United Kingdom currently illustrate how the same evidence can be used within different decision frameworks. The FDA authorized marketing of a specific deep-TMS system for OCD in 2018 based on device-level evidence and has since cleared additional systems for the indication. This allows clinical use of cleared devices according to their labeling in the United States.
By contrast, the NICE HealthTech guidance on rTMS for OCD states that safety raises no major concerns but that efficacy evidence is inadequate in quantity and quality, and recommends use only in the context of research. The guidance was migrated into NICE's current HealthTech format in January 2026 without changing that substantive recommendation.
The 2026 clinical-practice guideline update takes a third position: it recognizes rTMS as an augmentation option in resistant OCD and identifies several protocol families with supportive evidence, while emphasizing methodological limitations and the evolving evidence base. These are not three answers to an identical regulatory question. FDA clearance, NICE interventional-procedure guidance, and a psychiatric treatment guideline have different mandates, evidentiary frameworks, and thresholds for routine adoption.
How to evaluate a TMS clinic for OCD
A credible clinic should be able to identify the exact OCD protocol rather than merely advertise “TMS.” Ask which device and coil will be used, whether that exact device/protocol is FDA-cleared for OCD or being used off-label, which brain target is stimulated, the frequency and intensity, how motor threshold is determined, how many pulses and sessions are planned, and how targeting is performed.
Ask how the diagnosis was confirmed and how severity will be measured. A baseline Y-BOCS or comparable structured severity measure is useful, but it should sit inside a full clinical assessment. Ask what response threshold the clinic uses, when symptoms will be reassessed, whether functional outcomes are tracked, and what happens if improvement is partial or absent.
If the protocol uses symptom provocation, ask who designs it, how individualized it is, what level of distress is targeted, and how the clinic distinguishes provocation from ERP. A clinic should be able to explain why the provocation is included without claiming that its independent causal benefit is already settled science.
Ask how concurrent ERP and medication are coordinated. If a clinic tells patients to stop medication or psychotherapy simply because TMS is beginning, that recommendation should have a clear clinical rationale and appropriate prescribing oversight. The best treatment plan is integrated rather than organized around the marketing needs of one device.
Finally, ask about safety screening, hearing protection, seizure procedures, management of headaches or local pain, implant compatibility, clinician availability during treatment, and emergency protocols. The International OCD Foundation TMS overview is a useful patient-oriented reference for understanding established OCD TMS procedures and questions to discuss with a treatment team.
What the evidence supports — and what it does not
Established evidence: repeated TMS can reduce OCD symptom severity more than sham stimulation on average, and a specific high-frequency deep-TMS protocol targeting the mPFC/ACC network has randomized multicenter evidence and U.S. FDA clearance. Multiple meta-analyses support a treatment signal across several targets and frequencies. Short-term tolerability is generally favorable, with local discomfort and headache the most common adverse effects and seizure a rare but important risk.
Evidence with important uncertainty: the optimal target and frequency, how much symptom provocation adds beyond TMS itself, which patient characteristics predict response, how long benefits persist beyond the short follow-up periods used in many trials, and the best maintenance strategy. Network analyses can compare protocols statistically, but the underlying evidence is still composed largely of modest-sized trials with heterogeneous methods.
Claims that exceed current evidence: that TMS cures OCD, that one commercially available coil is proven superior for every patient, that symptom provocation has independently proven causal superiority, that a high real-world response percentage can be applied to every clinic, or that TMS makes first-line ERP and medication obsolete. The strongest clinical interpretation is more useful: TMS is a legitimate noninvasive treatment option with a real but variable effect, best used with protocol-level precision inside a comprehensive OCD treatment plan.
Frequently asked questions about TMS for OCD
Is TMS a recognized treatment for OCD?
Yes. Specific TMS systems and protocols are FDA-cleared for OCD in the United States, and randomized trials and meta-analyses support efficacy. Its place in routine care varies across guidelines and health systems. The current evidence most strongly supports TMS as an adjunct or advanced option after established treatments have been adequately considered.
Is deep TMS better than regular TMS for OCD?
Deep TMS has the best-known FDA-cleared OCD protocol and a pivotal multicenter randomized trial, but the broader rTMS literature also shows benefit at DLPFC and SMA/pre-SMA targets. Current evidence does not justify a universal claim that deeper stimulation is always superior. Device, target, frequency, intensity, protocol fidelity, and patient selection all matter.
How many TMS sessions are used for OCD?
The pivotal deep-TMS course was delivered on weekdays over six weeks, and FDA-cleared protocols commonly use a multiweek acute course. Exact session count and scheduling are device- and protocol-specific. A clinic should provide the labeled or evidence-based schedule it is using and explain any deviation.
Does TMS hurt?
TMS can be uncomfortable, especially at first. Common sensations include tapping or pressure on the scalp, facial or jaw muscle contractions, and headache. Most reported adverse effects in OCD trials have been mild or moderate and transient, but tolerability varies. Severe or unusual pain should be reported so the team can reassess positioning, intensity, and safety.
Can TMS cause a seizure?
Yes, seizure is a recognized rare risk of TMS. Risk is reduced by following established stimulation limits and screening for relevant medical, neurologic, medication, substance, and sleep-related factors. Anyone with a seizure history or a factor that may lower seizure threshold needs individualized assessment before treatment.
Can I take OCD medication during TMS?
Often, yes. Participants in the pivotal deep-TMS trial could remain on stable OCD medication. Medication changes should be coordinated with the prescriber because they can affect symptoms, interpretation of response, and sometimes seizure risk. TMS is frequently used as an adjunct rather than as a reason to discontinue medication.
Is symptom provocation the same as ERP?
No. TMS symptom provocation is a brief procedure intended to activate an OCD-relevant brain state around the time of stimulation. ERP is a comprehensive behavioral treatment involving planned exposure and prevention of rituals and safety behaviors across repeated learning experiences. They can be used in the same treatment plan but serve different functions.
How soon will I know whether TMS is working?
Response is normally assessed across several weeks. Some responders improve before the acute course is complete; others improve later or during short-term follow-up. Repeated standardized measurement is more reliable than judging the treatment from a single good or bad day.
Does TMS permanently cure OCD?
No evidence supports describing TMS as a permanent cure. Some patients experience clinically meaningful symptom reduction, and short-term follow-up suggests that benefits can persist after the acute course. Long-term durability and the best maintenance strategy remain less certain, and OCD often requires ongoing relapse-prevention and management planning.
Is TMS the same as magnetic therapy sold for home use?
No. Clinical TMS uses medical devices that generate precisely timed, high-intensity magnetic pulses according to defined stimulation parameters and requires professional targeting, calibration, monitoring, and safety procedures. Consumer magnets and generic “magnetic therapy” products are not equivalent to therapeutic TMS.
Can children or teenagers receive TMS for OCD?
The strongest OCD TMS evidence and U.S. clearances have historically focused on adults, and pediatric use requires specialist consideration of age-specific evidence, regulation, consent, neurodevelopment, and safety. For children and adolescents, established pediatric OCD treatments such as family-informed CBT with ERP and appropriately managed medication remain central. A pediatric specialist should evaluate any neuromodulation proposal.
What should I ask before paying for TMS for OCD?
Ask for the exact device, OCD indication, target, frequency, intensity, session count, evidence supporting that protocol, response definition, baseline and follow-up measurement plan, symptom-provocation procedure, safety screening, integration with ERP and medication, expected out-of-pocket cost, insurance requirements, and the plan for nonresponse or relapse. Specific answers are more informative than a clinic's overall TMS success percentage.
Bottom line
TMS is a real evidence-based neuromodulation option for OCD, particularly as an adjunct or advanced treatment when established care has not produced enough improvement. The evidence includes a positive multicenter sham-controlled deep-TMS trial, several supportive meta-analyses, FDA-cleared OCD systems in the United States, and increasingly detailed protocol comparisons. The signal is clinically meaningful for a subset of patients, but average effects are moderate and response varies substantially.
The most important question is therefore not simply “Does TMS work?” It is “Which TMS protocol, for which person, at what point in the OCD treatment pathway, measured by which outcomes, and integrated with which established treatments?” A high-quality answer requires correct diagnosis, protocol-level precision, realistic interpretation of effect sizes, careful safety screening, and continued attention to ERP, medication, and overall functional recovery.