"Did I actually say that?"
"Where exactly were my hands?"
"Did I look at them too long?"
"Maybe if I replay it one more time, I'll finally know."
That last sentence is the whole disorder in eleven words. Somewhere right now, someone is on their fourth hour of reconstructing a two-minute conversation — not because it was important, but because their mind flagged it, and now the tape is running again, frame by frame, in search of the one detail that would close the case.
That's mental reviewing. It's checking, performed on memory.
And because it happens entirely inside the skull, it gets missed — by the people doing it, and too often by clinicians who conclude the person "just has obsessions, no compulsions." That conclusion is almost always wrong. So-called Pure O is not obsession without compulsion; it's obsession with compulsions that happen to be invisible (Williams et al., 2011). Reviewing is one of the most common.
Reviewing promises retrospective certainty. Memory — as we'll get to — is structurally incapable of issuing it.
What the replay is actually doing
Ordinary remembering retrieves. Reviewing interrogates.
You're not recalling the party. You're deposing it:
Replaying the event to determine "what really happened."
Reconstructing the timeline — where you stood, what your hands did, the exact word order of a sentence.
Comparing this replay against the last replay, checking for consistency.
Scanning the memory for intent: "Did I want that to happen?"
Re-running the interaction from the other person's point of view to grade how you came across.
In the cognitive-behavioral model, this is a compulsion, full stop: a covert response performed to neutralize the distress of an obsession (Salkovskis, 1985). The tell is repetition plus purpose. You're interrogating the same material past the point where any new information could possibly exist.
Where the tape runs
Harm OCD. Replaying the drive home for the bump you might have felt. Reviewing the moment you held the knife while your partner walked past. "Was there a flicker of wanting?"
Real event and false memory OCD. The signature theme. A real memory — a party in college, a childhood game, a drunken night — gets reviewed hundreds of times to determine whether something terrible happened. Each replay produces a slightly different edit. The mind reads the edits as concealment. They're not. They're what memory reconstruction looks like.
Sexual orientation OCD. Reviewing past crushes and physical responses for "signs." Re-examining a glance at the gym like deposition testimony.
Relationship OCD. Replaying the early months to verify the feelings were real. "Was I actually happy at dinner, or was I performing?"
Scrupulosity. Re-running a confession to check whether it was complete. Reviewing a moment of anger for whether it constituted sin.
Contamination. Mentally retracing your path through the hospital lobby — what you touched, in what order, whether your hand went near your face.
Health OCD. Reconstructing when the headache started. "Was the sensation there yesterday, or am I remembering it wrong?"
Existential OCD. Replaying moments of feeling "unreal" to determine whether they prove something about your mind.
Different content. Same machine: retrieve, inspect, fail to certify, retrieve again.
Why the replay never settles anything
Memory does not work the way reviewing assumes it works.
Remembering is reconstructive. Every retrieval rebuilds the event from fragments plus current context, and every rebuild can introduce variation — this is established memory science, confirmed across decades including Loftus's work on memory malleability (Loftus, 2005). There is no master tape. When you review, you're comparing this retrieval to the last retrieval — checking an edit against an edit.
Reviewing a memory to obtain a fixed, verified, courtroom-grade record is like photographing a painting to prove the paint isn't wet. The method can't produce what's being demanded.
Worse: the method degrades the thing it's inspecting. Van den Hout and Kindt (2003) showed that repeated checking reduces memory vividness and confidence — and Radomsky and Alcolado (2010) extended this to mental checking specifically.
Read that again, because it's the whole trap:
Reviewing a memory makes you trust it less.
The doubt that launched the review gets stronger with every pass. This is why people with real event OCD report the memory feels less clear after a year of reviewing than the day it happened. A year of inspection did what inspection does to reconstructive material: wore it smooth.
Then add the reinforcement loop:
Replay → momentary "okay, I think nothing happened" → relief → the reviewing gets stronger.
Replay → no resolution → urgency → the next replay gets scheduled.
The compulsion wins either way. And every hour spent in the archive teaches your threat system that this memory is dangerous and must be resolved before life can continue (Rachman, 1997).
It can't be resolved. It was never going to be. That's not pessimism — it's how reconstructive memory works.
How ERP addresses it
ERP for reviewing has the same architecture as ERP for any ritual, adapted for one that lives in working memory.
Under the inhibitory learning model (Craske et al., 2014), the goal is not to review until anxiety fades. The goal is expectancy violation. The prediction being tested: "If I don't resolve this memory, I can't function."
So you don't resolve it. And you function. And your nervous system takes notes.
In practice:
-
Exposure to the unresolved version. An uncertainty script, written or recorded: "Maybe something happened at that party. Maybe I'll never know. I'm choosing not to check." Not a confession. Not a conclusion. Practice holding the open file without processing it.
-
Response prevention for retrieval. When the urge to replay arrives, you name it — "reviewing" — and put attention back on the task in front of you. Not suppression; that backfires for well-documented reasons covered in our article on thought suppression. The memory is allowed to exist. The investigation is declined.
-
Catching the smuggled review. Reviewing hides inside exposures — people write the uncertainty script while mentally checking it against the memory. Good ERP catches that.
Most clients who commit to this work report the memory loses its gravitational pull — it becomes a thing that happened (or didn't) rather than an open case file — though individual results vary.
How ACT addresses it
Reviewing treats a memory as a problem to be solved. ACT treats it as an experience to be had.
Defusion: "I'm having the thought that I need to figure out what happened." A thought. Not a subpoena.
Acceptance: letting the unresolved memory sit in awareness — ambiguity and all — without granting it a hearing.
Values: every hour in the archive is an hour not spent in the life the archive is supposedly protecting. The ACT question is not "what happened?"
It's "who do you want to be while not knowing?"
Twohig et al. (2018) support integrating ACT with exposure for OCD, and the fit here is natural.
What to practice instead
An unreviewed memory feels like an open wound at first. That discomfort is the price of exit — and it's temporary in a way the compulsion is not.
-
Name the session. The moment you notice you're back in the tape: "Reviewing." One word. Then attention out — the task, the room, the conversation.
-
Close the case formally. Pick the memory OCD keeps litigating and close it for 30 days. No new hearings. Every urge to reopen gets the same answer: "Case closed. Maybe I'll never know."
-
Practice the maybe. Once daily, say the uncertain version out loud: sixty seconds, flat delivery, no analyzing afterward. This is exposure, not journaling.
-
Catch the second-order review. After practicing, the mind will try to review how the practice went. Same compulsion, lab coat on. Decline it too.
-
Log time recovered. Track minutes spent reviewing each day. The falling number is more persuasive than any reassurance — and unlike reassurance, it's real data.
The past doesn't need your verification. It needs your absence from the archive.
Reviewing is one branch of a larger family — memory checking and rumination are its closest relatives. The full map is on our mental compulsions pillar, with the treatment model on our ERP therapy page and theme context on our OCD themes overview.
Work with a therapist who treats this directly
I provide online OCD therapy using ERP grounded in the inhibitory learning model, integrated with ACT, via telehealth in Texas, Washington, New Hampshire, and Florida. If intrusive memories are driving the reviewing, our guide to finding a therapist for intrusive thoughts covers what to look for. The first step is a free 15-minute consult call.
Felix Murad, M.Ed., LPC-S, LMHC, CMHC, NCC — Licensed Professional Counselor-Supervisor. Licensed by the Texas Behavioral Health Executive Council. Individual results vary; this article is educational and not a substitute for treatment.
References
Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour Research and Therapy, 58, 10–23.
Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year investigation of the malleability of memory. Learning & Memory, 12(4), 361–366.
Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35(9), 793–802.
Radomsky, A. S., & Alcolado, G. M. (2010). Don't even think about checking: Mental checking causes memory distrust. Journal of Behavior Therapy and Experimental Psychiatry, 41(4), 345–351.
Salkovskis, P. M. (1985). Obsessional-compulsive problems: A cognitive-behavioural analysis. Behaviour Research and Therapy, 23(5), 571–583.
Twohig, M. P., Abramowitz, J. S., Smith, B. M., Fabricant, L. E., Jacoby, R. J., Morrison, K. L., Bluett, E. J., Reuman, L., Blakey, S. M., & Ledermann, T. (2018). Adding acceptance and commitment therapy to exposure and response prevention for obsessive-compulsive disorder: A randomized controlled trial. Behaviour Research and Therapy, 108, 1–9.
van den Hout, M., & Kindt, M. (2003). Repeated checking causes memory distrust. Behaviour Research and Therapy, 41(3), 301–316.
Williams, M. T., Farris, S. G., Turkheimer, E., Pinto, A., Ozanick, K., Franklin, M. E., Liebowitz, M., Simpson, H. B., & Foa, E. B. (2011). Myth of the pure obsessional type in obsessive-compulsive disorder. Depression and Anxiety, 28(6), 495–500.
FAQ
Is mental reviewing the same as rumination?
They overlap but aren't identical. Reviewing targets the past — verifying what happened. Rumination is broader repetitive analysis aimed at meaning, causes, or the future. Both function as compulsions in OCD; the distinction matters mainly for designing response prevention.
Why does my memory get blurrier the more I review it?
Because repeated checking — including purely mental checking — reduces memory confidence and vividness (Radomsky & Alcolado, 2010; van den Hout & Kindt, 2003). The tool you're using to get certainty is the thing destroying it.
What if something really did happen?
ERP doesn't require concluding that nothing happened. It requires accepting that reviewing cannot deliver certainty either way, and that a meaningful life doesn't wait for a verdict. That stance holds whether the feared event is imagined, ambiguous, or real.
Does mental reviewing mean I have "Pure O"?
"Pure O" is a colloquial label for OCD whose compulsions are covert. Reviewing is a compulsion — so the presentation isn't compulsion-free. The compulsions are just internal.
Can this be treated over telehealth?
Yes. Mental rituals are especially well-suited to telehealth ERP, since the work happens in language and attention rather than physical space.