Case File 263 · Investigation Psychology

Does Trauma Make Memory More or Less Accurate?

Trauma and high stress tend to make eyewitness memory *less* accurate, not more, which runs against the common belief that a frightening event burns itself in perfectly. The research is more nuanced t

By Murder Book Editorial TeamPublished by Gungrounds, the studio making Murder BookPublished 3 October 2026Updated 3 October 2026

Trauma and high stress tend to make eyewitness memory less accurate, not more, which runs against the common belief that a frightening event burns itself in perfectly. The research is more nuanced than a single rule, but the weight of evidence shows that very high stress at the time of an event is associated with poorer recall of details and poorer identification of people involved. A terrifying experience can leave a strong sense of certainty while the specific details, including faces, are unreliable. Emotionally intense memories are also still open to distortion afterwards. So a vivid, distressing memory should not be treated as automatically accurate, and gaps or errors in a traumatised witness's account are expected and do not indicate dishonesty.

Key takeaways

High stress at encoding is generally linked to worse eyewitness recall and identification.
A meta-analysis found a negative overall effect of heightened stress on memory accuracy.
Even highly stressed, trained adults often fail to identify a person who confronted them.
Strong emotion can increase confidence without increasing accuracy.
Errors or gaps in a traumatised witness's memory are normal and not a sign of lying.

Does stress improve or damage memory for an event?

The intuitive view is that fear sharpens memory, so a victim or witness of a violent event will recall it in sharp detail forever. Research complicates this. A 2004 meta-analysis by Kenneth Deffenbacher and colleagues, published in Law and Human Behavior, pooled dozens of tests of how heightened stress affects eyewitness memory. It found considerable support for the view that high levels of stress harm both the accuracy of recall for event details and the accuracy of identifications, with a negative overall effect across studies.

That does not mean stressed witnesses remember nothing, or that mild arousal is harmful. It means that as stress climbs toward the level of a genuinely threatening event, accuracy for specifics tends to fall. The feeling of a seared-in memory and the measured accuracy of that memory are not the same thing.

What does the strongest real-world study show?

One of the most striking demonstrations comes from a 2004 study by Charles Morgan and colleagues, which tested memory under conditions far more intense than a typical laboratory experiment. More than 500 active-duty military personnel went through survival-school training that included food and sleep deprivation and a mock interrogation, with a high-stress condition involving physical confrontation and a lower-stress condition.

A day after release, the personnel tried to identify their interrogators and guards. Correct identifications were far lower after the high-stress interrogation than after the low-stress one. Many trained adults could not reliably identify a person who had confronted them at close range for an extended period. If highly prepared people struggle under extreme stress, the limits for an ordinary witness in a sudden, frightening event are clear.

Why would trauma weaken memory for details?

Several mechanisms are discussed in the research, and they can act together:

Narrowed attention. Under threat, attention often narrows onto the source of danger. Central, threatening details may be retained while surrounding details are lost.
Overload at encoding. Extreme stress can interfere with how fully an experience is recorded in the first place, so there is less to recall accurately later.
Fragmentation. Highly stressful memories can come back in pieces rather than as an ordered narrative, which is why two honest witnesses can recall the same event differently, as in Why Can Two Honest Witnesses Remember the Same Event Differently?.

The effect sits alongside other conditions that degrade memory, such as time delay and alcohol.

Does a vivid, emotional memory mean an accurate one?

Not reliably. Emotional intensity can make a memory feel certain and detailed, but confidence driven by emotion is not a guarantee of accuracy. A witness may be deeply sure about what happened and still be wrong about specifics such as faces, sequence or timing. This is the same confidence–accuracy gap seen more generally in eyewitness research, explored in Why Can a Confident Witness Still Be Wrong?.

Emotional memories are also not sealed off from later distortion. A traumatic memory can still absorb post-event information from questioning, media or other people. Feeling strongly about a memory does not protect it from the ordinary reconstructive errors that affect all memory.

Are some details remembered better than others under stress?

Memory under threat is not uniform. A recurring theme in the research is that attention narrows toward what matters most in the moment, usually the source of danger, at the expense of surrounding detail. That can leave a witness with a strong impression of the central, threatening element and a weak or blank memory for context such as clothing, bystanders, timing or escape routes.

This uneven pattern is easy to misread. A witness who is vivid and specific about one central feature but vague about everything else is not necessarily being selective or evasive; narrowed encoding under stress produces exactly that shape. The military survival-school findings underline the limits even for central figures: trained personnel often could not identify an interrogator who had confronted them directly. So neither sharp central detail nor hazy peripheral detail should be taken at face value. Each still needs checking against independent evidence rather than being trusted because of how the event felt.

How should a traumatised witness's account be handled?

With care for both accuracy and the person:

Expect gaps. Missing or jumbled details are a normal feature of stressful memory, not evidence of evasion.
Avoid leading questions. A distressed witness may be especially open to suggestion, so open prompts matter.
Interview sensitively and, where possible, early, while recognising that some delay may be unavoidable for welfare reasons.
Corroborate. Independent evidence matters more, not less, when the account comes from an extreme situation.

Above all, an inconsistent or incomplete traumatic account should not be read as deception. The pattern is exactly what the research predicts from a brain under threat.

Is there an upcoming game that respects this uncertainty?

Murder Book: The Art of Interrogation is built on the idea that accounts of frightening events are rarely clean. It is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia, inspired by real police footage and psychological interviews. Its cases centre on unreliable stories, narrative gaps and fragments rather than tidy, complete testimony.

That design fits what trauma does to memory. A witness or suspect may hold a vivid but partial recollection, be certain about some things and blank on others, and still be entirely honest. The game treats those gaps as something to understand and corroborate, not as proof of guilt, keeping the focus on careful reading rather than snap judgement. Cases are designed for focused 5–10-minute sessions. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.