Context-Dependent and State-Dependent Memory: Evidence and Limits
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Memory & Forgetting · Written by Artificial Sapiens
Context-dependent memory is the tendency for retrieval to improve when cues from the external environment at remembering resemble cues present during learning. State-dependent memory describes a related dependence on internal conditions, such as a physiological or affective state. Both concern the relationship between encoding and retrieval. Neither establishes that the original location or mood is necessary for a memory to exist, and neither provides a universal rule for studying or taking exams.
The central finding is conditional: reinstating relevant cues sometimes helps people recall what they learned, particularly on tasks requiring unaided recall. The size and reliability of the effect depend on the material, the cues, what the person attended to, the retrieval test, and alternative cues that remain available. A major meta-analysis of environmental-context studies, historical failures to replicate the effect, and more recent evidence of memory distortions all make those qualifications essential.
This guide explains external and internal retrieval contexts, the strongest experiments, successful and unsuccessful replications, current explanatory models, the practical implications for studying and everyday remembering, and the limits of claims about emotion, substances, eyewitness testimony, and brain mechanisms. For the broader framework, see how memory works and the specialist overview of retrieval cues and recognition.
What Is Context-Dependent Memory?
In an experimental context-dependent memory effect, researchers compare performance when the environment at retrieval matches the environment at encoding with performance when it differs. A benefit of the matching condition suggests that environmental features supplied retrieval information. Such features may include a room, background scene, sound, smell, physical setting, or another aspect of a situation. The term 'environmental context' often refers specifically to incidental background information that is not part of the item to be learned.
Imagine learning a list of words in a garden and later being asked to recall them either in that garden or indoors. If garden cues became associated with the learning episode, returning to the garden could help search memory for the words. Yet a person who can reconstruct the garden mentally, has strong semantic clues, or has rehearsed the words in many places may show little location advantage. This is why a plausible example must remain an illustration rather than a promise.
Context can be perceptual, spatial, semantic, temporal, social, or internally generated. Experimental studies vary in whether a background is incidental or deeply integrated with the material. That distinction matters: remembering a word printed on a particular background and remembering a factual concept learned in one classroom are not interchangeable research tasks. The Smith and Vela review specifically examined incidental environmental context.
External context versus internal state
External context refers to features of the situation, such as the room and its sensory cues. Internal state refers to conditions of the person, including experimentally manipulated mood, drug state, or other physiological circumstances. Their effects can interact. A familiar room might evoke a familiar emotional state, so an apparent location effect could partly reflect how the environment feels rather than its visual appearance. Eich's experiments on mood and place examined this possibility directly.
Context dependence is a relation, not a kind of memory
Context dependence describes an influence on access to information; it is not a separate memory store alongside episodic, semantic, or procedural memory. Episodic recollection may be sensitive to contextual details, while well-learned factual knowledge can be retrieved with powerful conceptual cues across many environments. The evidence is task-dependent, so it would be inaccurate to say that all episodic memory is context-bound or that semantic memory never benefits from context. See human memory systems and processes.
What Is State-Dependent Memory?
State-dependent memory occurs when a match between a person's internal state during encoding and retrieval facilitates memory relative to a state mismatch. The APA Dictionary of Psychology includes biological and psychological state in its definition. Studies have investigated alcohol or other pharmacological conditions and induced moods, but the results vary with the task and quality of the manipulation.
An important asymmetry often disappears in popular accounts: a relative matching-state benefit does not mean that an impaired state is a good state for learning. If intoxication weakens attention or encoding, learning and testing while intoxicated can remain inferior overall to learning and testing while sober. A matched impaired condition might outperform an impaired-to-sober switch on one recall measure without outperforming the sober-to-sober condition.
State-dependent, mood-dependent, and mood-congruent memory
Mood-dependent memory refers to better access when the mood at test resembles the mood at learning, irrespective of whether the target information is emotionally positive or negative. Mood-congruent memory is a different pattern: a current mood can favor material whose emotional meaning is similar, such as selecting negative memories while feeling sad. Mood dependence concerns a match across two moments; congruence concerns similarity between emotional content and current mood. They can co-occur without being the same mechanism.
A classroom illustration: a person studies neutral material while cheerful and remembers it later while cheerful. That is an example of mood dependence if a controlled comparison demonstrates the advantage. Recalling sad episodes more readily during sadness illustrates mood congruence. Real life includes additional factors—attention, expectations, arousal, rumination, and current concerns—making anecdotal inference weak. Eich's review of mood-dependent memory identified the nature of the material, the type of memory test, and the strength of mood induction as important moderators.
How Do Context and State Influence Retrieval?
Encoding specificity: a cue helps when it connects with learning
The classic encoding-specificity principle of Tulving and Thomson holds that a retrieval cue's effectiveness depends on how information was encoded, including the associations formed during learning. A seemingly weak cue can be useful when it was part of the original episode; a generally strong association can prove less useful if it does not connect to that particular encoding. The principle concerns compatibility between retrieval information and the earlier experience, not an exact duplication of every sensation.
When someone remembers a conversation after hearing a sound from the original event, that sound may narrow the set of candidate memories. In experiments, matching environments can sometimes serve this function even if the surroundings were incidental. The person also brings general knowledge, the question being asked, and other clues, so retrieved information emerges from several sources rather than from one environmental trigger.
Cue availability, diagnosticity, and competition
A cue becomes useful when it differentiates the target from competing memories. A classroom shared by hundreds of lessons can be a weak cue for one particular fact; a distinctive scent encountered during one event may be more discriminating. Reinstating a cue that is common to many experiences can also bring competing information to mind. The mere presence of the same cue at two times is therefore insufficient to predict benefit.
The question asked at retrieval matters as much as the surroundings. A prompt for the meaning of a concept may activate well-practiced semantic links; a prompt for where and when an episode occurred may recruit contextual detail. This helps explain why the same environmental manipulation may affect free recall but have little effect on recognition.
Mental context reinstatement
Physical return is one way to recover cues, but people can sometimes mentally reconstruct an earlier setting: picturing the room, the sequence of events, the conversation, or the route they took. This is a plausible route to retrieving related details when those details were encoded and remain discriminating. Mental reinstatement can also invite inference and confabulation when a person fills gaps using expectations. It therefore improves access in some settings without guaranteeing historical accuracy.
Context is partly reconstructed at retrieval
Recent evidence challenges the assumption that objects are always the remembered items and background scenes are always merely context. Across six experiments, Gardette and Bastin (2026) found that what functions as item or context depends on the retrieval question and stimulus properties. Their work supports a reconstructive view, although the laboratory design does not establish an all-purpose neural model or a universal effect in everyday situations.
The Classic Underwater Divers Experiment—and What Replication Found
In 1975, Godden and Baddeley studied divers learning word lists on land or underwater and then recalling those lists in a matching or mismatching environment. Their original results favored matched study-and-test environments in free recall. The finding became a widely taught demonstration because land and underwater conditions provide conspicuously different sets of cues.
The actual experimental claim was narrower than the textbook slogan 'study where you will be tested.' The participants learned relatively simple material under unusual conditions, and the outcome was free recall of words. Divers, deep-water logistics, auditory presentation, memory load, and environmental change are distinctive features of this paradigm. Results cannot be converted directly into an effect size for college examinations, clinical remembering, or professional performance.
In a subsequent 1980 study of recognition memory, Godden and Baddeley did not find a corresponding land–underwater context advantage for recognizing studied words among alternatives. The difference between free recall and recognition already suggested that explicit retrieval cues supplied by the test may reduce the influence of incidental environmental information.
Unsuccessful replication and the correction record
A 2021 replication by Jaap M. J. Murre tested 16 divers across land/underwater encoding and recall combinations and did not reproduce the predicted same-context advantage. This result does not erase the original experiment or show that context can never matter. It demonstrates that the specific effect is less mechanically repeatable than the familiar story implies.
The replication's statistical method was subsequently revisited. In a published addendum, Murre explained a repeated-measures analysis correction and reported that the reanalysis reinforced the conclusion that this replication had not succeeded. Including the addendum matters because evaluation of a famous experiment depends on corrected analyses as well as the original article.
Other challenges to the simple rule
An eight-experiment investigation by Fernández and Glenberg (1985) failed to find a reliable environmental matching advantage across their procedures, including a closer replication of a prior positive report. Differences in instructions, encoding conditions, accessibility of internal context, and retrieval support remain possible explanations. A failed effect in one method cannot be treated as proof of a zero effect in every method, but it places a clear limit on certainty.
The correct synthesis is that environmental context can be a retrieval cue under identifiable experimental circumstances, and the specific same-room advantage is variable. A large, complex literature needs to be evaluated through multiple designs, moderator analyses, and replications rather than one famous image of a diver.
What Do Reviews and Meta-Analyses Show?
The foundational Smith and Vela (2001) review and meta-analysis synthesized research on incidental environmental context-dependent memory. It supports the existence of environmental-context effects under some conditions and examines factors influencing their size. One important proposal is that attention directed toward noncontextual information during learning or testing can weaken dependence on ambient cues. The work is a review of a particular literature, not a global numerical guarantee for every everyday setting.
These studies differ in materials, the distinctiveness of environments, whether context is incidental or intentionally integrated, the time between learning and test, and whether outcomes are free recall or recognition. Pooling effects requires attention to such differences. Older studies also vary in sample size and reporting practice; newer direct replications can reveal fragile patterns. No honest synthesis turns the literature into a rule that students must occupy the exam room to remember.
The available evidence is much stronger for the general possibility of cue-dependent retrieval than for an exact amount of improvement from changing rooms, a guaranteed effect of particular moods, or a standard prescription for all learners. Research demonstrations explain mechanisms while educational recommendations must additionally prove practical transfer.
Does Context Affect Free Recall, Cued Recall, and Recognition Equally?
Free recall asks a person to produce studied material with few explicit reminders. Recognition supplies candidate items and asks whether they were encountered. Cued recall supplies some information linked to the target. When tests give strong item-specific cues, the incremental value of incidental environmental features may be reduced. The original diver recall finding and the null recognition follow-up provide a concrete comparison, though they do not establish a universal law of testing.
Another complication is that matching context can affect the decision process used in recognition. A familiar background may increase the feeling that an object belongs to a studied episode, even when that object is new but similar. Thus some newer recognition experiments reveal effects of context that involve both accurate hits and errors. The outcome measured—hits, false alarms, discrimination, confidence, or free-recall quantity—must be stated before judging a 'memory improvement.'
Context can increase false recognition
In experiments with objects superimposed on backgrounds, Doss, Picart, and Gallo (2018) found that reinstating context increased correct recognition of studied objects and also false recognition of similar new objects. Their result shows why a context-related increase in familiarity or 'remembering' responses is not automatically an increase in accuracy.
In three experiments, Bayoud, Pierce, and Gallo (2026) extended this pattern to associated-word false-memory tasks: reinstated context could increase both true and false recognition and reported recollective experience. A stronger warning to avoid misleading context weakened and destabilized both effects in one experiment. These findings are experimentally important, while the degree of transfer to courtroom testimony or ordinary disputes remains a separate question.
Internal-State Evidence: Alcohol, Mood, Arousal, and Physiological Conditions
Alcohol-induced state dependence
In the classic Goodwin and colleagues (1969) experiment, volunteers completed learning tasks while sober or under alcohol and were tested later in matching or mismatching states. The investigators found state-related differences for some recall and interference measures but not a significant matching-state influence on recognition. The design is historically important, yet older laboratory effects cannot be used as a recommendation to drink before learning or remembering.
Alcohol impairs judgment, attention, and other functions and introduces substantial safety risks. A relative retrieval match does not restore details never adequately encoded or make impaired testimony dependable. Deliberately reproducing intoxication to remember something is a poor and potentially dangerous application of these findings. The relevant scientific claim is about experimentally observed state matching under constrained conditions.
Mood-dependent memory
Mood research is more complicated than an easy happy-while-learning/happy-while-testing slogan. Eich (1995) reviewed factors associated with detecting mood dependence, including whether material was generated internally, the type of memory measure, and whether mood shifts were sufficiently strong. Across studies, induction methods and participant characteristics complicate interpretation.
A change in mood may alter attention, the associations formed at encoding, the way questions are interpreted, or access to particular autobiographical material. If a person remembers an upsetting experience when distressed, the observation alone cannot tell whether a state-match mechanism operated. A controlled test would need to separate memory strength, emotional congruence, motivation, repetition, and the cues supplied by the present situation.
Arousal, stress, sleepiness, caffeine, and pain
These states can influence cognition through multiple pathways, including changes in alertness and attention. Such effects should not all be called 'state-dependent memory.' To establish state dependence specifically, a study must compare matched and mismatched internal states at encoding and retrieval while accounting for direct effects on learning and test performance. The existence of some experimental drug- or mood-state effects does not justify assuming equivalent results for caffeine, exhaustion, stress, or pain.
The practical priority is a safe and functional condition for comprehension and learning, rather than reproducing an uncomfortable or risky state. Sleep disorders, trauma-related symptoms, dissociation, intoxication complications, and neurocognitive disease have their own clinical literatures and should not be reduced to this general retrieval principle.
What Is Known About the Brain Mechanisms?
Memory research associates the hippocampus and connected medial temporal systems with relational and episodic representations of events and places. Neuroscience investigations also study how elements of a previously encountered setting are represented and reinstated during remembering. Such evidence provides plausible mechanisms for cue-dependent recollection, but it is a separate inferential step to claim that a particular laboratory environmental-context benefit is caused by one uniquely identified brain circuit.
For example, Masís-Obando, Norman, and Baldassano (2026) examined how stable neural representations of spatial contexts can support reinstatement of subsequently placed objects. The study advances a mechanistic account of spatial scaffolding; its findings should not be generalized automatically to everyday mood matching, educational exam performance, or all forms of context dependence.
Models emphasizing pattern completion, hippocampal binding, cue overlap, predictive inference, and reconstructive retrieval can be complementary at different levels. The evidence does not warrant a simplistic picture in which a room or emotion locks one memory inside a single neural 'file.' For the general processes, see encoding, storage, and retrieval.
Everyday Examples: What They Illustrate and What They Do Not Prove
Walking back into a room to remember an intention
If you enter a room and forget why, returning to the earlier location can supply cues linked to the intention. This is a sensible everyday strategy because the earlier scene or action sequence may prompt a retrieval route. It does not prove that a context-dependent effect caused that particular lapse; attention shifts, divided goals, interference, and incomplete initial encoding are alternative explanations.
Remembering a conversation at a familiar location
Revisiting a café may prompt recollection of a conversation that took place there. The environment can provide distinctive cues, but memories are reconstructive and may combine accurate details with later assumptions. Familiarity of place should not be mistaken for independent verification of what someone said.
Music, odors, and personal memories
A particular song or scent may become associated with an experience and later cue personal recollection. Here the cue may be unusually salient, emotionally meaningful, or embedded in the event itself; it differs from the incidental-room manipulations in much of the laboratory literature. Strong subjective vividness is compatible with both genuine recollection and inaccurate details. Personal episodes are also shaped by repeated storytelling and subsequent experience.
Studying in a library and testing in a classroom
A matching environment could sometimes provide small retrieval support, but most exam settings present questions, concepts, and recognizable material that already act as retrieval cues. Preparation that works only in one room is vulnerable to changes. A robust approach is to learn concepts well, practice retrieving them without notes, and test them in more than one ordinary setting, while keeping adequate conditions for concentration.
Can Context Matching Help Students Remember More?
The strongest practical advice is to make target knowledge accessible with multiple retrieval routes. When an environment will be predictable, studying there may feel familiar and reduce distractions, but the research does not establish that matching the exact classroom consistently improves examination performance. It is often impossible to recreate an exam room, and most educational materials are more complex than laboratory word lists.
In Smith's 1982 experiments, learning across multiple environmental contexts supported later free recall under particular conditions compared with studying in one room; the effect was not found on every memory test. These findings motivate testing more than one cue arrangement rather than tying learning to a single setting. They are limited laboratory evidence, not a guaranteed study intervention for every curriculum.
For lasting retention, combining comprehension with repeated effortful retrieval and distributed practice has a stronger direct educational evidence base than attempting to recreate a room or emotional state. Roediger and Karpicke (2006) demonstrated benefits of retrieval testing for delayed recall of prose in controlled studies; readers can also explore spaced repetition and the spacing effect. This comparison concerns relative evidence for realistic strategies, not an exact head-to-head effect size for every course.
A practical, evidence-aligned study routine
First, learn and explain the material in your own words so that meaning connects concepts. Second, close your notes and attempt retrieval from a question, problem, or blank page; correct errors using feedback. Third, return after an interval and repeat. Fourth, vary ordinary conditions modestly—a desk, library, and quiet classroom—while maintaining appropriate lighting, safety, and concentration. Fifth, practice with exam-like questions because the retrieval demands themselves matter.
Recreating safe elements of the expected test situation, such as timing or answering without notes, may also help familiarity and test preparation. The evidence-based rationale for practicing exam-like tasks includes transfer-appropriate processing and task familiarity; it should not be presented as proven environmental context dependence. Avoid deliberate sleep loss, alcohol, or distress induction in pursuit of state matching.
What about using the same scent, playlist, or outfit?
A distinctive cue can theoretically become associated with learning, but consumer claims that wearing the same perfume, playing the same song, or dressing alike will reliably unlock exam answers go beyond the available evidence. Songs may distract, scents may be disallowed or affect others, and clothing may have no relevant learned association. If a harmless cue is personally useful, treat it as an optional reminder rather than a substitute for understanding and practice.
Can Changing Context Make Memory More Flexible?
Relying on one setting can sometimes leave a learner with fewer independent routes into the material. Retrieval across contexts may promote flexibility by repeatedly activating a target from varying prompts. This is an interpretation consistent with parts of the experimental literature, but generalizable educational benefits need direct demonstrations with meaningful outcomes and suitably controlled comparisons.
There is also no reason to force constant environmental change. A stable, quiet workplace can aid attention and reduce interruptions. Alternating a few workable settings is an option when convenient; depth of learning, retrieval practice, and sustainable habits usually deserve greater priority than environmental manipulation.
A useful question for study design is whether the learner can answer a conceptually equivalent question without depending on the page layout, a favorite desk, or the exact wording used during revision. Varied questions and retrieval prompts address this robustness directly.
When Context Reinstatement Is Misleading or Risky
The same cues that support retrieval can strengthen familiarity and false recognition. In the 2018 and 2026 studies of context reinstatement, increased endorsement did not reliably mean greater discrimination between old and new material. This matters when evaluating confident personal accounts and suggests care when asking people to reconstruct environments for highly consequential recollections.
Context reinstatement can be incorporated into some investigative interviewing procedures, but specific forensic techniques require their own evidence evaluation and safeguards against suggestion. A reader cannot infer from these memory experiments that every vivid recovered detail is accurate, that revisiting a scene proves a claim, or that a witness should be encouraged to generate additional details without careful attention to source and uncertainty.
State-matching arguments are particularly risky when applied to distress, substances, or clinical care. There is no scientific basis here for advising someone to reproduce trauma-related distress, intoxication, or another harmful state as a memory aid. Questions about abuse, dissociation, flashbacks, or therapy should be addressed within their specific clinical and forensic literatures, with professional standards rather than mnemonic folklore.
What the Evidence Supports—and What Remains Uncertain
Well supported at the level of cognitive principle
Retrieval depends on cues and the relation between current information and learning. Some well-controlled experiments document advantages from reinstating environmental or internal-state information. Task demands, cue quality, and encoding conditions help determine whether effects are observed. The encoding-specificity framework and environmental review/meta-analysis are foundational for this conclusion.
Mixed or conditional evidence
The size and reproducibility of ambient same-room benefits vary; classic findings have both positive results and direct failures to replicate. Recognition can be less sensitive than free recall to incidental context, although some recognition paradigms demonstrate important context effects. Mood-state experiments have identifiable moderators, and findings from pharmacological conditions do not automatically generalize to ordinary emotions.
Claims that exceed available evidence
There is no general proof that taking an exam in the same room, reproducing the same mood, or using a signature scent ensures superior memory. Matching an intoxicated state is not a recommended practice. Context reinstatement does not guarantee accurate details, as false-recognition studies demonstrate. Current brain research supports investigations of memory context representation but does not supply a single settled explanation for all behavioral effects.
Methodological priorities for future research
Stronger evidence will require preregistered and adequately powered studies, explicit separation of incidental versus integral cues, direct comparisons of free recall and recognition, objective measures of hits and false alarms, and tests with meaningful educational or everyday outcomes. The distinction between effects on subjective familiarity and accurate information recovery is especially important after recent research.
Context Dependence Versus Other Memory Concepts
Encoding specificity describes a general relationship between learning and retrieval cues, whereas context dependence denotes an observed pattern involving matched contextual information. Retrieval practice is an intervention that makes information more accessible through repeated remembering; it is not equivalent to being in the same room. Environmental-context effects are distinct from interference, which involves competition among memories, although competing traces can influence cue usefulness.
A memory can feel familiar because an earlier context has returned without yielding detailed, verifiable recollection of a particular event. Source monitoring concerns how people determine where information came from. Reconstructive and false-memory research examines how present cues, expectations, and later information influence what a person reports. These areas overlap causally, but each has its own primary research question.
State dependence is also not a diagnosis, proof of repression, or evidence that a person has inaccessible memories that can only be recovered in a particular mental state. Most ordinary retrieval failures are best investigated through attention, encoding, competing demands, cue availability, health context, and time course rather than through speculative hidden-state explanations.
Frequently Asked Questions
What is a simple example of context-dependent memory?
A person may remember something learned in a particular room more easily after returning to that room because sights and spatial cues from the original episode provide retrieval support. The example illustrates a possibility; the effect is not guaranteed for every memory or location.
What is an example of state-dependent memory?
A laboratory participant might recall some material better when tested in the same induced physiological condition in which it was learned than after a state change. Such a matching advantage must be measured against other conditions; it does not show that the induced condition improves memory overall.
What is the difference between state-dependent and context-dependent memory?
Context-dependent memory typically refers to a match in external cues such as settings or background scenes. State-dependent memory refers to a match in internal conditions such as mood or drug state. Both involve relations between encoding and retrieval, and both can interact.
Does context-dependent memory actually work?
Yes, in the sense that contextual retrieval effects have been demonstrated and analyzed scientifically. The same-context advantage is conditional and sometimes fails to replicate; it is not a dependable shortcut for every test. Smith and Vela (2001) synthesize the environmental literature.
Why did divers remember words better underwater in the classic experiment?
The original 1975 results were consistent with environmental cues assisting free recall when learning and test locations matched. A recognition follow-up did not reproduce the same advantage, and a later direct recall replication was unsuccessful. The best explanation therefore treats the classic finding as one task-dependent result rather than a fixed law.
Can I study in my exam room to get better grades?
A familiar exam environment could have practical comfort or cue advantages, but scientific evidence does not guarantee higher grades from matching rooms. Reliable methods include understanding the material, spaced retrieval practice, and practicing question types that resemble the real assessment.
Should I study in the same mood as the exam?
A comfortable, alert state is preferable to deliberately provoking anxiety, sadness, intoxication, or exhaustion. Mood-dependent memory is a conditional experimental finding, and recreating an unhealthy state has no established role as a safe study strategy.
Is mood-congruent memory the same as mood-dependent memory?
No. Mood congruence refers to a mood favoring emotionally similar content; mood dependence refers to a match between mood at learning and mood at recall. Researchers study the two phenomena separately.
Does context matter more for recall or recognition?
Incidental environmental context has often shown a clearer influence on free recall than on recognition, where test items supply strong cues. However, some recognition paradigms show context-related increases in both correct and false recognition. The answer depends on the task and outcome.
Can a smell or song trigger a memory?
Yes, a sensory cue associated with an experience can prompt recollection. The association may be personally distinctive and emotionally significant. Feeling that a memory is vivid does not by itself validate every detail.
Does revisiting the place where something happened make an eyewitness more accurate?
Returning to a setting may revive cues, but experiments show that context can also increase false endorsements. Accuracy requires corroboration and careful interviewing; recollected details should not be assumed accurate solely because the setting feels familiar.
Can context-dependent memory explain forgetting a name or walking into a room and forgetting why?
Contextual cues may help recover an intention or a name, but ordinary lapses have multiple explanations, including attention, weak encoding, interference, and delays. One anecdote cannot isolate an experimental context-dependent effect.
Is state-dependent memory the same as dissociative amnesia?
No. State-dependent retrieval is a general experimental memory phenomenon. Dissociative amnesia is a clinical concept with separate diagnostic considerations. Matching-state experiments neither establish a diagnosis nor demonstrate that apparently missing autobiographical memories can be recovered reliably by inducing a past state.
Can a different context make a memory stronger?
Studying or retrieving in varied settings can sometimes diversify available cues, as suggested by particular experiments, but not every study or test shows a benefit. The stronger transferable recommendation is varied meaningful retrieval practice under healthy conditions.
Is context-dependent memory stored in the hippocampus?
The hippocampus and connected systems contribute to contextual and relational episodic memory, and current studies examine neural context representations. Assigning every context-dependent behavioral effect to a single structure oversimplifies a distributed, task-sensitive system.
What is the safest practical takeaway?
Use relevant cues when trying to remember, and strengthen learning through comprehension, repeated retrieval, and spaced practice. Context can sometimes help, but it should remain one source of information among many, with special caution whenever accuracy, safety, or clinical interpretation matters.
Related Articles
For foundational concepts, read Human Memory: Types, Processes, and How It Works, How Memory Works: Encoding, Storage, and Retrieval, and Memory Retrieval: Recall, Recognition, and Cues. For study methods and memory-system differences, see Spaced Repetition: How the Spacing Effect Supports Long-Term Memory and Short-Term Memory vs. Long-Term Memory: Key Differences.
References
American Psychological Association. (2018). State-dependent memory. APA Dictionary of Psychology. Official definition.
Bayoud, A., Pierce, B. H., & Gallo, D. A. (2026). Context reinstatement increases true and false recognition of associated words: Evidence from the Deese/Roediger-McDermott (DRM) illusion. Memory & Cognition. https://doi.org/10.3758/s13421-026-01892-4.
Doss, M. K., Picart, J. K., & Gallo, D. A. (2018). The dark side of context: Context reinstatement can distort memory. Psychological Science, 29, 914–925. https://doi.org/10.1177/0956797617749534.
Eich, E. (1995a). Mood as a mediator of place dependent memory. Journal of Experimental Psychology: General, 124, 293–308. https://doi.org/10.1037/0096-3445.124.3.293.
Eich, E. (1995b). Searching for mood dependent memory. Psychological Science, 6, 67–75. https://doi.org/10.1111/j.1467-9280.1995.tb00309.x.
Fernández, A., & Glenberg, A. M. (1985). Changing environmental context does not reliably affect memory. Memory & Cognition, 13, 333–345. https://doi.org/10.3758/BF03202501.
Gardette, J., & Bastin, C. (2026). Context in memory is reconstructed based on retrieval questions. Communications Psychology. https://doi.org/10.1038/s44271-026-00504-w.
Godden, D. R., & Baddeley, A. D. (1975). Context-dependent memory in two natural environments: On land and underwater. British Journal of Psychology, 66, 325–331. https://doi.org/10.1111/j.2044-8295.1975.tb01468.x.
Godden, D., & Baddeley, A. (1980). When does context influence recognition memory? British Journal of Psychology, 71, 99–104. https://doi.org/10.1111/j.2044-8295.1980.tb02735.x.
Goodwin, D. W., Powell, B., Bremer, D., Hoine, H., & Stern, J. (1969). Alcohol and recall: State-dependent effects in man. Science, 163, 1358–1360. https://doi.org/10.1126/science.163.3873.1358.
Masís-Obando, R., Norman, K. A., & Baldassano, C. (2026). Spatial contexts with reliable neural representations support reinstatement of subsequently placed objects. Nature Human Behaviour, 10, 164–181. Journal article.
Murre, J. M. J. (2021). The Godden and Baddeley (1975) experiment on context-dependent memory on land and underwater: A replication. Royal Society Open Science, 8, 200724. https://doi.org/10.1098/rsos.200724.
Murre, J. M. J. (2022). Addendum to: Murre (2021), The Godden and Baddeley (1975) experiment on context-dependent memory on land and underwater: A replication. Royal Society Open Science. Official full text.
Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17, 249–255. https://doi.org/10.1111/j.1467-9280.2006.01693.x.
Smith, S. M. (1982). Enhancement of recall using multiple environmental contexts during learning. Memory & Cognition, 10, 405–412. https://doi.org/10.3758/BF03197642.
Smith, S. M., & Vela, E. (2001). Environmental context-dependent memory: A review and meta-analysis. Psychonomic Bulletin & Review, 8, 203–220. https://doi.org/10.3758/BF03196157.
Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80, 352–373. https://doi.org/10.1037/h0020071.