Why Can't I Stop Checking, Even When I Know It's Fine?

You are in the car. You checked the lock. You remember checking the lock, or you remember doing something that was probably checking the lock, and the memory has the texture of a photograph of a photograph. So you go back. And the going back works, for about as long as it takes to reach the car again.

What follows is what research on obsessive compulsive disorder, or OCD, has actually established about this, including one finding that reorganises the whole problem.

How common is compulsive checking?

It is the most frequently reported compulsion there is. In the National Comorbidity Survey Replication, 79.3 percent of US adults with lifetime OCD reported checking, ahead of every other obsession or compulsion measured. A 2020 meta-analysis opens by stating flatly that compulsive checking is the most common ritual among people with OCD.

Checking also forms one of the symptom dimensions that replicate across countries, usually described as harm concerns with checking. OCD overall has a lifetime prevalence of around 2 to 3 percent, with a mean age of onset of 19.5 years in US survey data.

One more number, because it tends to land: a 2025 meta-analysis of 31 studies and nearly 6,000 patients found a pooled gap of about 80 months, roughly six and a half years, between the onset of OCD and seeking help.

Why does checking make the doubt worse?

Because repeated checking appears to damage how much you trust your own memory, while leaving the memory itself intact. This is the counterintuitive core of the research, and it was demonstrated directly.

In a now classic experiment, healthy participants checked a virtual gas stove on a first and last trial. In between, one group checked the stove 18 more times and the other checked light bulbs 18 times. Memory accuracy was unchanged in both groups. But in the group that had repeatedly checked the stove, the vividness and detail of the memory, and their confidence in it, had all declined.

That effect has since been replicated in people who actually have OCD. Participants repeatedly checked kitchen appliances, and both clinical and non-clinical participants reported significantly reduced memory confidence, vividness and detail afterwards. The effect appeared only for checking the item later recalled, not for irrelevant checking. The authors' reading is the important part: these declines are a consequence of checking rather than a cause of it.

So the sentence to take away is not that you have a bad memory. It is that going back to look is the thing degrading your sense of having looked.

Is it really about the house burning down?

Apparently less than everyone assumes. A meta-analysis of 22 studies comparing 663 people with OCD against 614 controls found that checking did not differ between threatening and neutral task conditions, which is hard to square with a purely threat-driven account. What did differ was perceptual decision-making: people with OCD checked significantly more on tasks involving their senses, with no such difference on reasoning tasks.

The authors concluded that the data support models built on distrust of the senses rather than on fear of harm. Put plainly: it may be less that you believe the stove will burn the house down and more that looking at the stove has stopped feeling like information.

That has a practical consequence. Arguing yourself out of the catastrophe, which is what most people try first, aims at the wrong target.

Why does reassurance stop working?

Because it works, briefly, which is exactly the trap. In a study of people with OCD, all participants said reassurance typically made them feel better, describing relief and a weight lifting. And all described the relief as short lived, one participant saying it wells back up and then it is like getting grabbed in the guts again.

Reliable short-term relief paired with unreliable long-term relief is the textbook shape of a behaviour that strengthens itself. The checking is not failing. It is succeeding at the wrong timescale.

Does knowing it makes no sense mean it is not OCD?

No, and this is worth hearing early. Retained insight is characteristic of OCD. A major review notes that most patients with OCD are keenly aware their compulsive symptoms are excessive and wish they had more control over them, and the National Institute of Mental Health says the same. The gap between what you know and what you then do is the condition, not a sign you are making it up.

What does treatment for OCD actually involve?

The first line psychological treatment is cognitive behavioural therapy built around exposure and response prevention, usually shortened to ERP. Exposure means deliberately confronting the situation or the thought. Response prevention means not performing the compulsion afterwards. It is built as a graded hierarchy, starting with what is manageable, and the International OCD Foundation describes a typical course as 12 to 20 sessions of about an hour, with practice between sessions. The American Psychiatric Association's practice guideline, which dates from 2007 and has not been revised, puts expert consensus at 13 to 20 weekly sessions.

One thing has changed in how good practitioners frame it. The old model treated the anxiety dropping during a session, called habituation, as the sign that it was working. Current thinking favours inhibitory learning, in which exposure builds new non-threat associations that compete with the old ones. On this account, as Ryan Jacoby and Jonathan Abramowitz put it, habituation of anxiety becomes a pleasant side effect of exposure rather than a requirement for success. The target is the prediction, not the comfort. For a checker, that looks less like calming down about the stove and more like discovering that what you expected to be unbearable was not.

Medication is part of care for many people with OCD and is managed by a prescriber. The Healing Effect provides counselling and does not prescribe.

How well does it work, honestly?

Well enough to be worth doing, and not as a cure. Both halves matter.

A meta-analysis of 36 randomised trials found a large benefit for CBT with ERP, an effect size of 0.74 against control conditions combined. The same analysis found that the eight trials without evidence of researcher allegiance showed essentially no effect, while those with it showed a large one, and that CBT with ERP did not outperform other active psychological treatments or adequately dosed medication by much. That is a real caveat and it belongs next to the headline number.

On the longer view, a five-year study of 213 adults with OCD found full remission in 16.9 percent and partial remission in 22.1 percent, with 59 percent of those who remitted later relapsing. Around 40 to 60 percent of people treated for OCD continue to have residual symptoms. There is one genuinely encouraging detail for readers of this article specifically: in that five-year study, people whose primary obsessions concerned over-responsibility for harm, which is the checking dimension, had a significantly better prognosis than other presentations.

The honest summary is that OCD is usually managed rather than ended, that checking is among the more treatable presentations, and that the gap between untreated and treated is large.

Where to start in Arizona

If checking is taking real time out of your day, or you are late to things, or you have started asking someone else to confirm the door for you, that is enough reason to talk to someone. You do not need to have a name for it first.

The Healing Effect works with clients across Arizona, in person and by telehealth for clients located in the state. You can get in touch through our contact form and tell us what you are dealing with in your own words.

And keep the one finding that changes how this feels. Going back to check is not a weakness of will that you have failed to overcome. It is a behaviour that reliably removes the certainty it promises, which is why doing more of it has never once been enough.

Why do I keep checking when I know everything is fine?

Experimental research suggests the checking itself is responsible. When people repeatedly check an item, their memory for it stays accurate but the memory becomes less vivid, less detailed and less trustworthy to them. The effect has been replicated in people with OCD, and researchers describe these declines in memory confidence as a consequence of checking rather than a cause of it.

How common is checking in OCD?

It is the most commonly reported compulsion. In the National Comorbidity Survey Replication, 79.3 percent of US adults with lifetime OCD reported checking, more than any other obsession or compulsion measured, and a 2020 meta-analysis describes compulsive checking as the most common ritual among people with OCD.

What is exposure and response prevention?

It is a form of cognitive behavioural therapy in which you deliberately face the situation or thought that triggers the urge, then do not perform the compulsion. It is built as a graded hierarchy starting with what is manageable, involves practice between sessions, and typically runs 12 to 20 sessions of about an hour according to the International OCD Foundation.

Does the anxiety have to go down for ERP to work?

Not according to current thinking. The field has moved from habituation, where anxiety dropping within a session was the marker of success, toward inhibitory learning, where the aim is to build new non-threat associations. On that account, anxiety reduction is a welcome side effect rather than a requirement, and the measure of progress is whether your predictions were violated.

Can OCD be cured?

It is better described as managed. A five-year study of 213 adults with OCD found full remission in 16.9 percent and partial remission in 22.1 percent, with 59 percent of those who remitted later relapsing, and 40 to 60 percent of treated patients continue to have residual symptoms. That said, the same study found that presentations centred on over-responsibility for harm, which includes checking, had a significantly better prognosis than others.

The Healing Effect is not a crisis service. OCD carries a substantially elevated risk of suicide, which is one more reason not to wait years before telling someone. If you are in crisis or thinking about harming yourself, call or text 988 to reach the Suicide and Crisis Lifeline, available 24 hours a day, 7 days a week.

Sources:

Ruscio AM, Stein DJ, Chiu WT, Kessler RC. The epidemiology of obsessive-compulsive disorder in the National Comorbidity Survey Replication. Molecular Psychiatry. 2010;15(1):53 to 63 (checking endorsed by 79.3 percent of adults with lifetime OCD, the most common of all types measured; mean age of onset 19.5 years). Survey fielded 2001 to 2003 using DSM-IV criteria.

Strauss AY, Fradkin I, McNally RJ, Linkovski O, Anholt GE, Huppert JD. Why check? A meta-analysis of checking in obsessive-compulsive disorder: Threat vs. distrust of senses. Clinical Psychology Review. 2020;75:101807 (checking as the most common ritual; 22 studies, 663 OCD participants and 614 controls; no difference between threatening and neutral conditions; more checking on perceptual but not reasoning tasks; support for distrust of senses over threat-based models).

van den Hout MA, Kindt M. Repeated checking causes memory distrust. Behaviour Research and Therapy. 2003;41(3):301 to 316 (virtual gas stove paradigm; relevant versus irrelevant checking; accuracy unchanged while vividness, detail and memory confidence declined). We read the experimental detail and result as reported in the introduction of Giele CL, Engelhard IM, van den Hout MA, et al. Repeated checking induces uncertainty about future threat. Journal of Experimental Psychopathology. 2015;6(2):126 to 137, co-authored by van den Hout, rather than at the 2003 paper itself.

Radomsky AS, Dugas MJ, Alcolado GM, Lavoie SL. When more is less: Doubt, repetition, memory, metamemory, and compulsive checking in OCD. Behaviour Research and Therapy. 2014;59:30 to 39 (replication in a clinical OCD sample; reduced memory confidence, vividness and detail after relevant checking only; memory declines as a consequence of checking rather than a cause).

Halldorsson B, Salkovskis PM. Why do people with OCD and health anxiety seek reassurance excessively? An investigation of differences and similarities in function. Cognitive Therapy and Research. 2017;41:301 to 322 (all OCD participants reporting that reassurance made them feel better emotionally; the effects typically short lived; participant descriptions quoted).

Stein DJ, Costa DLC, Lochner C, et al. Obsessive-compulsive disorder. Nature Reviews Disease Primers. 2019;5:52 (lifetime prevalence of 2 to 3 percent; cross-culturally replicated symptom dimensions including harm concerns with checking; most patients keenly aware their symptoms are excessive; approximately half of patients failing to fully respond to a first-line treatment).

Pellegrini L, Giobelli S, Burato S, et al. Meta-analysis of age at help-seeking and duration of untreated illness in obsessive-compulsive disorder. Journal of Affective Disorders. 2025;380:212 to 225 (31 studies, 5,960 patients; pooled duration of untreated illness of 80.23 months).

National Institute of Mental Health. Obsessive-Compulsive Disorder: When Unwanted Thoughts or Repetitive Behaviors Take Over. NIH Publication No. 23-MH-4676, revised 2023 (ERP effective in reducing compulsive behaviours; many adults with OCD recognise their compulsive behaviours do not make sense; medication prescribed by health care providers).

International OCD Foundation. Exposure and Response Prevention (ERP), OCD Treatment Guide (situational and imaginal exposure; resisting the urge to engage in compulsions; graded hierarchy; a typical course of 12 to 20 sessions of about one hour with practice between sessions).

American Psychiatric Association. Practice Guideline for the Treatment of Patients With Obsessive-Compulsive Disorder. 2007, reaffirmed 2012 (expert consensus of 13 to 20 weekly sessions; booster sessions after successful ERP). Note this guideline dates from 2007 and has not been revised.

Jacoby RJ, Abramowitz JS. Inhibitory learning approaches to exposure therapy: A critical review and translation to obsessive-compulsive disorder. Clinical Psychology Review. 2016;49:28 to 40 (inhibitory learning rather than habituation; successful response to exposure occurring in the absence of habituation; habituation of anxiety as a pleasant side effect rather than a requirement for success; expectancy violation as the target).

Reid JE, Laws KR, Drummond L, Vismara M, Grancini B, Mpavaenda D, Fineberg NA. Cognitive behavioural therapy with exposure and response prevention in the treatment of obsessive-compulsive disorder: A systematic review and meta-analysis of randomised controlled trials. Comprehensive Psychiatry. 2021;106:152223 (36 trials, 2,020 participants; Hedges g 0.74 against all controls; g 0.95 where researcher allegiance was suspected against g 0.02 in the 8 trials without it; no significant advantage over other active psychological treatments or adequately dosed pharmacotherapy).

Eisen JL, Sibrava NJ, Boisseau CL, et al. Five-Year Course of Obsessive-Compulsive Disorder: Predictors of Remission and Relapse. Journal of Clinical Psychiatry. 2013;74(3):233 to 239 (213 adults; full remission 16.9 percent and partial remission 22.1 percent over five years; 59 percent of those who remitted subsequently relapsing; better prognosis for primary obsessions concerning over-responsibility for harm).

Nezgovorova V, Reid J, Fineberg NA, Hollander E. Optimizing first line treatments for adults with OCD. Comprehensive Psychiatry. 2022;115:152305 (approximately 40 to 60 percent of patients with OCD still exhibiting residual symptoms).

Fernandez de la Cruz L, Rydell M, Runeson B, et al. Suicide in obsessive-compulsive disorder: a population-based study of 36,788 Swedish patients. Molecular Psychiatry. 2017;22:1626 to 1632 (substantially elevated risk of death by suicide and suicide attempt compared with matched general population controls, persisting after adjustment for psychiatric comorbidity).