SSRIs are not considered addictive in the same sense as drugs associated with compulsive reward-seeking, but they can cause physiologic adaptation and discontinuation symptoms.

Medication for Anxiety Disorders: Options, Benefits, Risks, and Monitoring

Creator: Ukrainian Psychological Hub
Published: 2026-10-01

Oct 1

25 min read

Updated: Oct 8

Author: Ukrainian Psychological Hub · Published: October 1, 2026 · Editorial Policy

Medication can be an effective part of treatment for some anxiety disorders, but “anxiety medication” is not one treatment for one condition. Anxiety is a normal human state as well as a symptom that can occur in many medical and mental health conditions, while an anxiety disorder is a clinical diagnosis defined by a persistent pattern of fear or anxiety, avoidance, distress, and impairment. The World Health Organization ICD-11 Clinical Descriptions and Diagnostic Requirements classifies generalized anxiety disorder, panic disorder, agoraphobia, specific phobia, social anxiety disorder, separation anxiety disorder, and selective mutism within its anxiety or fear-related disorders grouping in ICD-11. Their medication evidence is not interchangeable.

For adults with generalized anxiety disorder, panic disorder, and social anxiety disorder, selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) have substantial evidence and are widely used. Other medications have narrower roles. Buspirone is principally an option for generalized anxiety disorder. Pregabalin is an evidence-based option for generalized anxiety disorder in some jurisdictions. Benzodiazepines can reduce anxiety rapidly but carry clinically important risks of physical dependence, withdrawal, sedation, misuse, and dangerous interactions. Beta blockers may reduce selected physical symptoms in performance situations, but a 2025 systematic review and meta-analysis found insufficient robust evidence for beta blockers as treatments for anxiety disorders.

Diagnosis matters. Medication evidence is strongest for some disorders and sparse for others. Exposure-based psychotherapy remains the central evidence-based treatment for specific phobia. Pediatric medication evidence for separation anxiety often comes from trials that enrolled mixed groups of children with separation anxiety disorder, generalized anxiety disorder, and social anxiety disorder. Selective mutism has only a small medication evidence base. A panic attack is an episode that can occur inside or outside panic disorder, and “anxiety attack” is an informal expression rather than a formal diagnosis. A screening score cannot determine which medication a person needs.

This article explains medication classes, disorder-specific evidence, likely benefits, adverse effects, dependence and withdrawal, monitoring, special populations, and the limits of medication. It does not provide an individual prescription, dose, medication combination, or taper. Those decisions require a qualified prescriber who can review diagnosis, medical history, other medications, substance use, pregnancy or breastfeeding, previous treatment response, and personal preferences.

What Medication for Anxiety Disorders Actually Means

The phrase “medication for anxiety” can refer to very different clinical targets: persistent generalized worry, recurrent unexpected panic attacks, fear of social scrutiny, avoidance of specific objects or situations, distress around separation, or a child who speaks normally in some settings but is consistently unable to speak in others. Those targets belong to different diagnostic pathways. Medication should be selected for the condition being treated and the person’s medical context, not simply because the word “anxiety” appears in the symptom description.

The World Federation of Societies of Biological Psychiatry guideline reviewed more than a thousand randomized controlled trials across anxiety and related disorders and identifies SSRIs and SNRIs as major first-line pharmacological options. That broad conclusion is useful, but it does not erase disorder-specific differences. NICE guidance for adult social anxiety disorder places individual disorder-specific CBT ahead of medication, while NICE guidance for adult generalized anxiety disorder recommends an SSRI when drug treatment is chosen. For specific phobia, the evidence center of gravity is exposure therapy rather than medication.

The useful medication questions are therefore: What diagnosis is being treated? What evidence exists for this medication in that diagnosis and age group? Is the use approved in this jurisdiction or off label? What benefits matter to the person? What harms, withdrawal effects, and interactions are relevant? How will response and safety be monitored?

Medication Does Not Turn Normal Anxiety Into a Disorder

Fear can be an adaptive response to immediate threat. Anxiety can be an anticipatory response to uncertainty. Worry can be a cognitive process that occurs with or without a disorder. Stress responses can follow demanding circumstances. None of these experiences automatically establishes an anxiety disorder or a need for medication.

Clinical treatment becomes relevant when a defined syndrome is present with sufficient persistence, intensity, distress, avoidance, or functional impairment. Medication may also be appropriate when anxiety symptoms occur within another diagnosed condition, but then the treatment target may be that condition rather than an anxiety disorder. OCD, PTSD, illness anxiety disorder, attachment anxiety, neuroticism, and historical “anxiety neurosis” are not simply subtypes of contemporary anxiety disorders.

The same principle applies to panic. A panic attack is a defined episode of intense fear or discomfort with rapid symptom escalation. It can occur in panic disorder, other mental disorders, medical conditions, or substance-related states. Treating “panic attacks” without assessing why they occur can miss the actual diagnosis.

The Main Medication Classes

SSRIs

Selective serotonin reuptake inhibitors are antidepressants that also treat several anxiety disorders. Their classification as antidepressants does not mean a person must have depression to benefit from them. Common examples include escitalopram, sertraline, paroxetine, and fluoxetine. Regulatory indications differ by drug, diagnosis, age, and country.

SSRIs act gradually rather than as immediate rescue medications. A 2024 meta-analysis of weekly response trajectories across 122 placebo-controlled trials found that SSRIs and SNRIs improved anxiety compared with placebo, while benzodiazepines produced faster early improvement. By around eight weeks, average response differences among the major classes were much smaller. This supports realistic expectations: early days do not reliably predict the final result, and meaningful benefit often develops over several weeks.

Common SSRI adverse effects can include nausea, diarrhea, headache, sleep disturbance, activation or jitteriness, sexual dysfunction, sweating, and appetite changes. A systematic review and network meta-analysis of 80 studies and 21,338 participants found adverse events were more common with SSRIs and SNRIs than placebo overall, with nausea the most frequent reported event. Individual drugs differed in tolerability patterns.

SNRIs

Serotonin-norepinephrine reuptake inhibitors, especially venlafaxine and duloxetine, also have strong evidence in selected anxiety disorders. Venlafaxine extended release has substantial evidence across generalized anxiety disorder, social anxiety disorder, and panic disorder; duloxetine has strong evidence for generalized anxiety disorder. SNRIs share many adverse effects with SSRIs and may also affect blood pressure or heart rate in some people, which can make cardiovascular monitoring relevant.

SSRIs and SNRIs are often grouped together because both are long-term antidepressant-class treatments with substantial anxiety-disorder evidence. They are not clinically identical. Previous response, discontinuation burden, interactions, co-occurring pain, sexual adverse effects, gastrointestinal effects, blood pressure, and patient preference can all matter when choosing between them.

Benzodiazepines

Benzodiazepines, including alprazolam, clonazepam, diazepam, and lorazepam, can reduce anxiety rapidly through GABA-A receptor effects. Their speed is real: in the 2024 response-trajectory meta-analysis, benzodiazepines improved symptoms faster than SSRIs or SNRIs in the first week. Rapid relief, however, is only one dimension of treatment.

The U.S. Food and Drug Administration requires boxed warnings for risks including abuse, misuse, addiction, physical dependence, and withdrawal. Physical dependence can develop during prescribed use and is not the same as addiction. Abrupt discontinuation or overly rapid dose reduction can cause serious withdrawal, including seizures. Combining benzodiazepines with opioids, alcohol, or other central nervous system depressants can substantially increase the risk of respiratory depression, overdose, and death.

Guidelines therefore restrict routine long-term benzodiazepine use. NICE advises against routine benzodiazepine treatment for generalized anxiety disorder except as a short-term measure during crises and advises against benzodiazepines for long-term management of panic disorder. The clinical question is not whether benzodiazepines can reduce anxiety—they can—but whether the expected benefit justifies dependence, withdrawal, sedation, cognitive, driving, fall, and interaction risks in a particular situation.

Buspirone

Buspirone is a non-benzodiazepine anxiolytic whose established role is mainly generalized anxiety disorder. It is taken regularly rather than as an immediate rescue medication. The current U.S. buspirone labeling is based on studies in outpatients whose clinical presentation corresponded closely to generalized anxiety disorder. It does not provide a basis for assuming equal efficacy in panic disorder, social anxiety disorder, specific phobia, or selective mutism.

Buspirone does not carry the same classic benzodiazepine dependence profile, but it still has adverse effects and interactions. Dizziness, nausea, headache, nervousness, and lightheadedness can occur. It also does not work immediately. A person switching from a benzodiazepine to buspirone may still need a separate benzodiazepine taper because buspirone does not prevent benzodiazepine withdrawal.

Pregabalin

Pregabalin is a gabapentinoid with evidence for generalized anxiety disorder. Regulatory status is jurisdiction-specific. The European Medicines Agency includes generalized anxiety disorder in adults among Lyrica’s indications, while the current U.S. Lyrica prescribing information does not list GAD. Describing pregabalin simply as “approved for anxiety” without naming the jurisdiction is inaccurate.

Pregabalin can cause dizziness, sleepiness, edema, and weight gain, and use requires attention to renal function. Misuse and physical dependence concerns are relevant for some patients. Its evidence is strongest in GAD; it should not be presented as a universal anxiety-disorder medication.

Hydroxyzine

Hydroxyzine is a sedating antihistamine with anxiolytic effects. A Cochrane review found some evidence of benefit in generalized anxiety disorder but judged the evidence too limited and biased to support it as a reliable first-line treatment. Current U.S. hydroxyzine labeling notes that effectiveness as an antianxiety agent for long-term use beyond four months has not been systematically assessed.

Hydroxyzine can cause sedation and anticholinergic effects and may be problematic in some cardiac rhythm contexts. Its role is therefore different from a maintenance SSRI or SNRI, and “fast-acting” should not be mistaken for “best long-term treatment.”

Beta Blockers

Beta blockers such as propranolol can blunt peripheral adrenergic symptoms such as tremor and rapid heartbeat. They are sometimes used off label for circumscribed performance situations. This is different from treating the full syndrome of social anxiety disorder, panic disorder, or generalized anxiety disorder. The 2025 systematic review and meta-analysis by Archer and colleagues found a lack of robust evidence that beta blockers are effective treatments for anxiety disorders. They also have medical contraindications and can affect heart rate and blood pressure.

Tricyclic Antidepressants and Monoamine Oxidase Inhibitors

Older antidepressants can be effective in selected anxiety disorders. Tricyclic antidepressants such as imipramine have evidence for panic disorder, and monoamine oxidase inhibitors have evidence in panic and social anxiety. Their use is generally more specialized because of adverse effects, overdose toxicity, dietary or drug interaction burdens, and monitoring requirements. In a 2022 network meta-analysis of panic disorder, tricyclic antidepressants, benzodiazepines, monoamine oxidase inhibitors, SSRIs, and SNRIs all showed higher remission rates than placebo, but almost every included study had some concern or high risk of bias. Efficacy ranking does not by itself determine the safest first choice.

Antipsychotics and Other Augmentation Strategies

Antipsychotics are not routine first-line treatments for primary anxiety disorders. NICE specifically advises against antipsychotic treatment for GAD in primary care. Augmentation strategies may be considered by specialists after an adequate diagnostic review and evidence-based treatment trials, but that is a different clinical problem from choosing an initial medication for an uncomplicated anxiety disorder.

Evidence by Diagnosis

One of the most common errors in online anxiety content is to take evidence from generalized anxiety disorder and silently apply it to every condition that contains fear or anxiety. The evidence is more uneven.

Generalized Anxiety Disorder

GAD has one of the strongest medication evidence bases. A 2025 Cochrane review of 37 randomized trials found high-certainty evidence that antidepressants improved treatment response compared with placebo in adults with moderate-to-severe GAD. The estimated number needed to treat for one additional beneficial response was about seven. Overall dropout rates were similar, while discontinuation because of adverse effects was more common with antidepressants.

A large 2019 Lancet network meta-analysis also found several pharmacological options effective, including duloxetine, pregabalin, venlafaxine, and escitalopram, with differences in acceptability. For a disorder-specific review, see Medication for Generalized Anxiety Disorder: Options, Benefits, and Risks.

Panic Disorder

Panic disorder is defined by recurrent unexpected panic attacks plus persistent concern, worry, or behavior change related to further attacks. A panic attack by itself is not panic disorder. The 2022 BMJ network meta-analysis found evidence that several medication classes increase remission relative to placebo, including SSRIs, SNRIs, benzodiazepines, tricyclic antidepressants, and monoamine oxidase inhibitors. The high risk-of-bias profile of the underlying trials and the safety differences between classes are crucial when interpreting those results.

For longer-term treatment, antidepressants are commonly favored because benzodiazepines carry dependence and withdrawal risks. Panic-focused CBT and interoceptive exposure are also major evidence-based treatments. See Medication for Panic Disorder: Options, Benefits, and Risks and the broader Panic Disorder guide.

Agoraphobia

Agoraphobia is its own diagnosis and can occur with or without panic disorder. Much medication research historically studied “panic disorder with or without agoraphobia,” which makes it easy to overstate medication evidence for standalone agoraphobia. Antidepressant treatment may be clinically reasonable when panic disorder or another medication-responsive anxiety disorder co-occurs, but exposure-based CBT remains a central treatment for agoraphobic avoidance. See Agoraphobia: Symptoms, Causes, Diagnosis, and Treatment.

Social Anxiety Disorder

Social anxiety disorder has both psychotherapy and pharmacological evidence. A Lancet Psychiatry network meta-analysis of psychological and pharmacological interventions found strong effects for individual CBT and evidence of benefit for SSRIs and SNRIs. NICE recommends individual disorder-specific CBT for adults as the initial treatment and, when a person prefers medication, recommends an SSRI such as escitalopram or sertraline.

Beta blockers deserve a separate boundary. They can reduce tremor or tachycardia in some performance settings but do not have robust evidence as treatments for the full disorder. For the disorder-specific medication page, see Medication for Social Anxiety Disorder: Options, Benefits, and Risks.

Specific Phobia

Specific phobia is the clearest example of why “all anxiety disorders use the same medication” is misleading. Exposure therapy is the preferred treatment. A 2022 meta-analysis found large effects for both single-session and multi-session exposure treatments, while a systematic review of exposure therapy identified a large evidence base focused on optimizing exposure success. Medication is not established as routine primary treatment for specific phobia.

Medication can occasionally be used for a separate co-occurring condition or in narrowly defined procedural contexts, but that should not be confused with treating the phobia itself. The popular Greek-derived name of a fear does not create a separate medication indication or a new DSM/ICD diagnosis.

Separation Anxiety Disorder

Separation anxiety disorder occurs in both children and adults, but much medication evidence comes from pediatric trials that grouped several anxiety diagnoses together. In the landmark Child/Adolescent Anxiety Multimodal Study, 488 young people with separation anxiety disorder, GAD, or social phobia were randomized to CBT, sertraline, their combination, or placebo. All active treatments outperformed placebo, with the combination producing the highest acute response rate. That trial supports sertraline for pediatric anxiety in a mixed diagnostic sample; it does not prove identical effect sizes for every individual disorder.

The AACAP clinical practice guideline concludes that CBT and SSRIs have considerable empirical support for child and adolescent anxiety disorders, with some support for SNRIs. Family involvement, developmental context, school functioning, and careful safety monitoring are especially important. See Separation Anxiety Disorder Treatment: CBT, Family Support, and Medication.

Selective Mutism

Selective mutism is primarily treated with behavioral and CBT-based approaches coordinated across home, school, and clinical settings. Medication evidence is much smaller than for GAD, panic disorder, or social anxiety disorder. An older double-blind placebo-controlled fluoxetine trial randomized only 15 placebo nonresponders; parent ratings favored fluoxetine on some outcomes, while clinician and teacher ratings did not show significant group differences and most children remained symptomatic. A later five-case controlled single-case study also suggested improvement but remained very small.

This is best described as limited, preliminary medication evidence. SSRIs may be considered as an adjunct in severe or persistent cases, particularly with substantial co-occurring anxiety, but the evidence base does not justify presenting medication as a universal first-line treatment for selective mutism. See Selective Mutism: Symptoms, Causes, Diagnosis, and Treatment.

How Much Can Medication Help?

Clinical trials report averages, not guarantees. Some people achieve remission, others obtain partial improvement, and some experience little benefit or cannot tolerate the medication. A broad 2026 network meta-analysis found that many anxiolytic drugs outperform placebo on anxiety outcomes, but its pooled sample was heavily shaped by generalized-anxiety and Hamilton Anxiety Rating Scale studies. It should not be read as a universal ranking for every anxiety disorder.

Benefit can mean fewer panic attacks, less persistent worry, lower anticipatory anxiety, better sleep, reduced physiological arousal, greater ability to enter feared situations, or improved work, school, and social functioning. Symptom reduction is useful, but functional recovery matters too. A medication that lowers a rating-scale score while causing intolerable sedation, sexual dysfunction, gastrointestinal effects, cognitive problems, or withdrawal difficulties may not be an acceptable treatment for that individual.

“Best medication” is therefore not a fixed league table. The appropriate option depends on the actual diagnosis, strength of evidence for that diagnosis, prior response, adverse-effect vulnerability, other conditions and medications, age, pregnancy or breastfeeding, substance-use history, practical access, and patient preference.

How Fast Do Anxiety Medications Work?

Time course varies by class. Benzodiazepines can produce noticeable relief within hours and showed faster average early improvement in the 2024 trajectory meta-analysis. Hydroxyzine can also act quickly as a sedating antihistamine. SSRIs, SNRIs, and buspirone usually require repeated use and improve symptoms gradually over weeks.

Early side effects can appear before therapeutic benefit. A person can experience nausea, sleep disturbance, or activation during the first days of an SSRI or SNRI and conclude that the treatment is “making anxiety worse.” Sometimes early activation is transient; sometimes it is clinically significant and requires review. Treatment should be reassessed rather than pushed through automatically.

A short trial can also be misleading. Clinicians usually allow an adequate trial at a tolerated regimen before judging efficacy unless adverse effects, emerging suicidality, mania, severe agitation, allergic reaction, or another safety concern requires earlier action. Exact timing and dose decisions belong to the prescriber because labels, age groups, diagnoses, and patient factors differ.

Side Effects and Tolerability

Adverse effects are part of the treatment decision. The 2023 systematic review and network meta-analysis of SSRI/SNRI adverse events found that participants receiving medication reported at least one adverse event more often than placebo participants, and medications differed in gastrointestinal, autonomic, sleep, and overall tolerability profiles.

Common SSRI and SNRI problems include nausea, diarrhea or constipation, headache, insomnia or sleepiness, sweating, sexual dysfunction, appetite changes, and early activation. SNRIs can also raise blood pressure in some patients. Tricyclic antidepressants can cause anticholinergic effects, orthostatic hypotension, cardiac conduction concerns, and greater toxicity in overdose. MAOIs carry major food and drug interaction constraints. Hydroxyzine can sedate and has anticholinergic and cardiac-rhythm considerations. Pregabalin can cause dizziness, somnolence, edema, and weight gain. Benzodiazepines can cause sedation, impaired coordination, memory problems, falls, and driving impairment in addition to dependence and withdrawal.

Side effects can resemble anxiety. Restlessness, tremor, insomnia, palpitations, dizziness, nausea, or sweating may come from the disorder, the medication, another drug, caffeine or other substances, withdrawal, or a medical condition. Similarity of symptoms does not establish the cause.

Dependence, Addiction, Tolerance, Withdrawal, and Discontinuation

Physical dependence means the body has adapted to a drug and withdrawal can occur after stopping or reducing it. Addiction involves compulsive use despite harm, craving, loss of control, or related behavioral features. A person can be physically dependent without having an addiction.

Tolerance means a given effect becomes smaller with repeated exposure. Withdrawal is the syndrome that occurs when a physiologically adapted system loses the drug too quickly. Discontinuation symptoms are the term often used for symptoms after stopping antidepressants and can include dizziness, flu-like sensations, sleep disturbance, irritability, anxiety, sensory symptoms, and gastrointestinal problems.

Benzodiazepines require the greatest dependence emphasis. The FDA boxed warning update states that physical dependence can occur even with prescribed use and that abrupt stopping or rapid reduction can trigger serious and potentially life-threatening withdrawal. Antidepressants are not considered addictive in the same reward-seeking sense, but they can cause physiologic adaptation and clinically significant discontinuation symptoms.

Withdrawal can mimic relapse. A sudden increase in anxiety after a rapid medication reduction may represent discontinuation, return of the underlying disorder, or both. Timing, symptom pattern, medication half-life, treatment duration, and the course after dose changes help a clinician differentiate them. This is one reason self-directed abrupt stopping is risky.

Monitoring Before Starting Medication

Good medication treatment starts with assessment, not a prescription pad. Before starting a drug, a clinician typically clarifies the target diagnosis and symptoms, severity and functional impairment, prior psychotherapy and medication history, medical conditions, pregnancy or breastfeeding, current medications and supplements, alcohol and drug use, allergies, and patient preferences.

Several additional checks can change the plan. A history of mania or hypomania matters because antidepressant treatment can be problematic in bipolar-spectrum illness without appropriate evaluation. Current fluoxetine prescribing information explicitly instructs clinicians to screen for bipolar disorder and monitor for mania or hypomania. Kidney function matters for pregabalin. Blood pressure and cardiovascular history can matter for SNRIs, beta blockers, tricyclic antidepressants, and selected other drugs. Fall risk and cognitive vulnerability matter in older adults. Opioid or sedative use is highly relevant before benzodiazepines.

Diagnosis should be revisited when the presentation is atypical. Palpitations, tremor, shortness of breath, dizziness, fainting, chest discomfort, insomnia, and agitation can occur in anxiety disorders, but they can also arise from arrhythmias, thyroid disease, asthma, anemia, medication effects, stimulant use, withdrawal states, or other medical conditions. A symptom overlap is not enough to label the cause “anxiety.”

Monitoring During the First Weeks

Early follow-up should ask two separate questions: Is it helping? Is it causing harm? Monitoring includes target symptoms, functioning, sleep, activation, gastrointestinal effects, sexual adverse effects, blood pressure when relevant, adherence, new medications or substances, and any signs of misuse or escalating sedative use.

Young people require particular attention to behavioral and mood changes after starting antidepressants. Current fluoxetine labeling carries a boxed warning about increased risk of suicidal thoughts and behaviors in children, adolescents, and young adults in short-term studies and calls for monitoring for worsening or emergent suicidal thoughts and behaviors. That warning does not mean antidepressants inevitably cause suicidality or should never be used in young people. The AACAP guideline supports SSRIs as evidence-based short-term treatment for pediatric anxiety disorders when clinically appropriate.

NICE adult GAD guidance recommends reviewing efficacy and side effects every two to four weeks during the first three months and every three months thereafter, with closer early monitoring for younger adults when starting SSRIs or SNRIs. Exact schedules vary by diagnosis, medication, age, health system, and individual risk.

Long-Term Monitoring

Once symptoms improve, monitoring changes from “Does it start working?” to “Is benefit sustained, and is the treatment still worth its burden?” Clinicians review symptom recurrence, functional recovery, adherence, weight or metabolic changes where relevant, blood pressure, sexual effects, sleep, cognition, substance use, and whether continued medication still fits the person’s goals.

Stopping immediately after improvement can increase relapse risk in some disorders and can produce withdrawal with some medications. Maintenance duration depends on diagnosis, number and severity of episodes, residual symptoms, comorbidity, prior relapse after discontinuation, adverse effects, and patient preference. Discontinuation should be planned rather than triggered solely by a calendar date or a single low symptom score.

What if the First Medication Does Not Work?

“Treatment failure” should not be declared until the underlying situation has been reviewed. The first questions are whether the diagnosis is correct, whether the target symptoms are the ones the medication is expected to treat, whether the trial was adequate, whether the medication was taken consistently, whether adverse effects limited treatment, and whether alcohol, stimulants, cannabis, sedatives, or another medication are complicating the picture.

Depending on the diagnosis, a clinician may consider switching within a class, switching classes, moving to disorder-specific psychotherapy, combining medication with psychotherapy, or using a second-line option. The sequence should follow disorder-specific evidence. Buspirone’s GAD evidence does not automatically make it the next step for panic disorder. A beta blocker’s ability to reduce tremor does not make it a treatment for generalized worry. Pregabalin’s GAD evidence does not establish equal benefit for selective mutism or specific phobia.

Repeated medication changes without diagnostic review can also miss a different problem, including bipolar disorder, substance-induced anxiety, medication-induced anxiety, PTSD, OCD, illness anxiety, sleep disorders, endocrine disease, arrhythmia, or another medical cause. These conditions can coexist with anxiety, but they are not interchangeable subtypes of “anxiety disorder.”

Medication and Psychotherapy

Medication and psychotherapy can both be evidence-based, but their relative place varies by disorder. For GAD, panic disorder, and social anxiety disorder, both medication and CBT have substantial evidence. For specific phobia, exposure therapy is the dominant treatment. For selective mutism, behavioral treatment across settings is foundational and medication evidence is limited.

Combination treatment is not automatically superior in every anxiety disorder. In the CAMS pediatric trial, CBT plus sertraline produced higher acute response than either alone in a mixed sample of separation anxiety, GAD, and social phobia. That finding is important, but it should not be generalized to every adult anxiety disorder or every medication-therapy pair.

Psychotherapy can address mechanisms that medication does not directly train, such as avoidance, safety behaviors, catastrophic interpretation, feared-situation learning, and behavioral flexibility. Medication can make symptoms more manageable; therapy can change learned patterns that maintain fear and avoidance. Treatment planning can use either, both, or a staged sequence based on diagnosis, preference, availability, and prior response.

Children and Adolescents

Medication decisions in children and adolescents require developmental assessment, family involvement, age-appropriate monitoring, and attention to school functioning. The AACAP guideline supports CBT and SSRIs as short-term evidence-based treatments for pediatric anxiety disorders and describes some support for SNRIs. A 2023 systematic review and meta-analysis likewise found SSRIs and SNRIs effective for pediatric anxiety disorders while noting increased behavioral activation.

Much of the pediatric evidence comes from mixed-diagnosis studies rather than isolated trials for every disorder. Regulatory approval also differs by drug, age, and diagnosis. A medication can have evidence in pediatric anxiety without holding a U.S. indication for the specific disorder being treated. Off-label prescribing can be evidence-based, but “off label” and “unsupported” are not synonyms.

Behavioral activation, agitation, sleep changes, irritability, and suicidal thoughts or behaviors deserve careful early monitoring. Families should know what changes warrant prompt contact with the prescriber and what symptoms require urgent help.

Pregnancy and Breastfeeding

Pregnancy does not reduce medication decisions to “safe” versus “unsafe.” Untreated severe anxiety can also affect health, functioning, sleep, nutrition, substance use, prenatal care, and postpartum wellbeing. The American College of Obstetricians and Gynecologists clinical practice guideline addresses psychiatric medication treatment during pregnancy and lactation and frames decisions around the risks and benefits of treatment in the context of the condition being treated.

A person who becomes pregnant while taking an anxiety medication should not abruptly stop it simply because of pregnancy. The next step is a medication-specific review of the evidence, gestational stage, prior relapse history, severity of illness, alternatives, and breastfeeding plans. Benzodiazepines, antidepressants, pregabalin, hydroxyzine, and other drugs each have different reproductive safety profiles; a broad “anxiety medicine in pregnancy” answer is not sufficiently precise.

Older Adults

Older adults may have different pharmacokinetics, medical comorbidities, fall risk, cognitive vulnerability, and polypharmacy. A 2025 Lancet Psychiatry systematic review and meta-analysis found evidence that antidepressants reduce anxiety symptoms in older adults, while evidence for benzodiazepine efficacy and safety was weak.

The 2023 AGS Beers Criteria generally list benzodiazepines as potentially inappropriate in older adults because of increased sensitivity and risks such as cognitive impairment, delirium, falls, fractures, and motor-vehicle crashes, while recognizing that exceptions can exist in selected clinical situations. The criteria are a decision-support tool rather than an automatic ban.

Start-up effects, orthostatic symptoms, sodium abnormalities with some antidepressants, kidney function, drug-drug interactions, and sedative load may all matter more with age. Medication review should include nonpsychiatric prescriptions and over-the-counter products as well as psychiatric medications.

Alcohol, Cannabis, Opioids, Stimulants, and Other Drugs

Substances can change both anxiety symptoms and medication safety. Alcohol may temporarily reduce tension and then worsen sleep, rebound anxiety, and sedation. Combining alcohol with benzodiazepines or other sedatives can be dangerous. Opioids plus benzodiazepines carry a particularly serious respiratory-depression and overdose risk highlighted by the FDA.

Cannabis can reduce anxiety for some people in some moments and provoke anxiety, panic, cognitive effects, or dependence in others; it is not a predictable substitute for disorder-specific treatment. Stimulants, high caffeine intake, decongestants, thyroid hormone excess, corticosteroids, and other medications can contribute to anxiety-like symptoms in susceptible individuals. A complete medication and substance list can prevent a “treatment-resistant anxiety” label from obscuring a drug-induced problem.

When Anxiety-Like Symptoms Need Urgent Medical Assessment

New or severe physical symptoms should not be diagnosed as anxiety online merely because they resemble panic. Chest pain, fainting, severe or worsening shortness of breath, a new irregular heartbeat, new focal neurological symptoms, seizure-like activity, severe allergic symptoms, suspected overdose, severe medication reaction, or rapidly worsening confusion require medical assessment based on the clinical situation.

Likewise, new suicidal intent, an inability to stay safe, severe agitation with loss of control, or symptoms suggesting mania or psychosis warrant urgent mental health assessment. Medication side effects, withdrawal, the underlying illness, substances, and independent medical conditions can overlap. The purpose of urgent assessment is to identify the cause rather than assume it.

Screening Scores Do Not Choose Medication

Questionnaires such as the GAD-7 measure symptoms and can help with screening or tracking severity. They do not establish every anxiety diagnosis, rule out medical causes, or select a medication. A person with a high score may have GAD, another anxiety disorder, depression with anxiety, trauma-related symptoms, substance effects, medication effects, a medical condition, or several overlapping problems.

Formal diagnosis requires a clinical assessment of symptom pattern, duration, triggers, avoidance, impairment, differential diagnoses, and context. Medication choice comes after that assessment. A lower score during treatment can be useful evidence of improvement, but it is one piece of outcome monitoring rather than proof that treatment should automatically be continued, changed, or stopped.

How to Discuss Medication With a Prescriber

A productive medication discussion focuses on tradeoffs rather than asking for a universally best drug. Useful questions include the following.

• Which diagnosis are we treating, and what makes that diagnosis the best fit?

• What evidence supports this medication for this specific disorder and age group?

• Is the use approved in this jurisdiction or off label?

• What symptoms and functional goals should improve if it works?

• How long should an adequate trial take before we judge response?

• Which early side effects are expected, and which require prompt review?

• Are there interactions with my other prescriptions, supplements, alcohol, cannabis, or other substances?

• Does this medication carry withdrawal or dependence risk?

• What monitoring is needed for blood pressure, kidney function, falls, mood activation, suicidality, or other risks?

• If it works, how will we decide how long to continue it and how to stop it safely later?

Shared decision-making is especially important because efficacy differences among reasonable options are often smaller than the differences in individual tolerability, interaction burden, withdrawal profile, and personal preference.

Frequently Asked Questions

What is the best medication for anxiety disorders?

There is no single best medication across all anxiety disorders. SSRIs and SNRIs have strong evidence for GAD, panic disorder, and social anxiety disorder, while other disorders have different treatment priorities. The best-supported option depends on diagnosis, age, medical context, previous response, adverse effects, interactions, and patient preference.

What is usually the first-line medication for anxiety disorders?

When medication is appropriate, SSRIs and SNRIs are the most broadly supported first-line classes across several major anxiety disorders. This is not equivalent to saying every SSRI or SNRI is first-line for every disorder. Specific phobia, selective mutism, and some developmental presentations have different evidence boundaries.

Which anxiety medication works fastest?

Benzodiazepines work faster than SSRIs and SNRIs on average, and hydroxyzine can also act quickly. Faster onset does not make a medication better for long-term treatment. Benzodiazepines carry dependence, withdrawal, sedation, and interaction risks.

How long do SSRIs and SNRIs take to work for anxiety?

Some improvement can begin within the first few weeks, while fuller benefit often takes several weeks. The time course differs by disorder, drug, person, and outcome. Early side effects can occur before benefit.

Can an SSRI or SNRI make anxiety worse at first?

Yes. Some people experience temporary activation, jitteriness, insomnia, agitation, or increased anxiety after starting or increasing an antidepressant. Severe, persistent, or concerning changes require clinical review rather than automatic continuation.

Are benzodiazepines addictive?

They can cause abuse, misuse, addiction, and physical dependence. Physical dependence can develop even when a benzodiazepine is taken exactly as prescribed. Dependence and addiction are related but distinct concepts.

Can I stop a benzodiazepine when I feel better?

Not abruptly if it has been used regularly. The FDA warns that sudden discontinuation or rapid dose reduction can cause serious withdrawal, including seizures. A taper should be individualized by the prescriber.

Are SSRIs addictive?

SSRIs are not considered addictive in the same sense as substances that produce compulsive reward-seeking, but they can cause physiologic adaptation and discontinuation symptoms. That is why gradual, planned discontinuation may be needed.

Does buspirone work for every anxiety disorder?

No. Buspirone’s established evidence and labeling are centered on generalized anxiety disorder. Evidence should not be automatically transferred to panic disorder, social anxiety disorder, specific phobia, separation anxiety disorder, or selective mutism.

Is pregabalin approved for anxiety?

Approval depends on jurisdiction. Pregabalin is authorized for generalized anxiety disorder in adults in the European Union, while the current U.S. Lyrica label does not list GAD. Its strongest anxiety evidence is for GAD.

Do beta blockers treat anxiety disorders?

They can reduce physical adrenergic symptoms such as tremor or rapid heartbeat and are sometimes used off label for performance situations. Current systematic-review evidence does not establish robust efficacy for treating anxiety disorders as a whole.

Is medication the main treatment for specific phobia?

No. Exposure therapy is the preferred evidence-based treatment for specific phobia. Medication may be used for a co-occurring condition or a narrow procedural circumstance, but it is not the standard primary treatment for the phobia itself.

Can medication treat selective mutism?

SSRIs have limited preliminary evidence, including small older fluoxetine studies, but behavioral and CBT-based interventions across school, home, and clinical settings remain foundational. Medication may be considered as an adjunct in selected severe or persistent cases.

Can a GAD-7 or anxiety test tell me which medication I need?

No. Screening scales measure symptoms or severity. They do not provide a formal diagnosis, rule out medical causes, or function as prescribing algorithms.

Is medication better than CBT?

That depends on the disorder and the person. Both are evidence-based for several major anxiety disorders. Individual CBT is a preferred initial treatment for adult social anxiety in NICE guidance, and exposure therapy is central for specific phobia. Medication can be highly useful for GAD, panic disorder, and social anxiety disorder. Combination treatment can help some people but is not automatically superior in every disorder.

Should I stop medication as soon as I feel better?

Not automatically. Stopping too early may increase relapse risk, and abrupt discontinuation can cause withdrawal with some drugs. Duration should be based on diagnosis, severity, residual symptoms, relapse history, adverse effects, and treatment goals.