Harm OCD is a form of obsessive-compulsive disorder in which the intrusive thoughts are about causing harm — that you might hurt someone you love, lose control with a knife in the kitchen, swerve the car into oncoming traffic, push someone onto the tracks. The thoughts arrive uninvited. They are repugnant to the person having them. And they are usually followed by hours of checking, avoiding, confessing and silently arguing back.
If that is what you are living with, here is the short answer: a thought of this kind is not a wish, a plan, or a warning. It is a recognised symptom of OCD, it is far more common than almost anyone realises, and it responds to a specific treatment. What follows is the fuller answer — including the one part that most articles on this subject leave out, because a purely reassuring version of this page would not be an honest one.
If you need help right now, get it from a person, not a page. If you are thinking about suicide, or if a thought of harming someone feels like something you want or intend rather than something you dread, contact someone now. In the US, call or text 988 (Suicide & Crisis Lifeline). In the UK and Ireland, call Samaritans on 116 123 — free, any time. Outside those regions, contact your local emergency number or crisis line.
What harm OCD actually is
“Harm OCD” is not a separate diagnosis. It is the everyday name for what clinicians call aggressive obsessions — one of the recognised symptom dimensions of OCD, sitting alongside contamination, symmetry, religious and sexual obsessions. The clinical definition is broad: intrusive thoughts of harming oneself or others, whether unintentionally or intentionally.
It is also, by a wide margin, not rare. A 2026 meta-analytic review in the Journal of Psychiatric Research screened 7,794 studies and coded 110 of them to estimate how common aggressive obsessions are among adults with clinician-diagnosed OCD. The lifetime prevalence was 70.3% (95% CI 61.6–77.7). Past-week prevalence was 52.6% (46.3–59.0). And for 28.0% of people with OCD, aggressive obsessions were their primary and most distressing symptom.
Read that again if you have spent years assuming you were the only one. The same authors describe aggressive obsessions as “experienced as distressing, commonly misdiagnosed by healthcare providers, and highly stigmatized” — a fair summary of why so few people say any of this out loud.
Intrusive thoughts themselves are not an OCD phenomenon at all. In a study that interviewed 777 university students at 15 sites in 13 countries across six continents, 93.6% reported at least one unwanted intrusive thought in the previous three months. Doubting intrusions were the most common category; repugnant ones the least. But almost nobody escapes them entirely.
So what turns a normal intrusion into OCD? Not the content. The psychologist Stanley Rachman set out the answer in a 1997 paper in Behaviour Research and Therapy that still underpins how the disorder is treated: obsessions are caused by “catastrophic misinterpretations of the significance of one’s thoughts,” and they persist for exactly as long as those misinterpretations do. Most people have the thought and discard it as noise. In OCD it is read as evidence — about what you might do, or about what you must secretly be — and everything that follows is an attempt to settle a question that cannot be settled.
This is not the same as wanting to
The word clinicians use is ego-dystonic: the thought is at odds with your values, and you experience it as alien and repellent. That is the opposite of intent, and it is the single most reliable feature distinguishing an obsession from a genuine urge.
On the question everyone actually wants answered, the clinical literature is unusually direct. In a 2009 review in Advances in Psychiatric Treatment, David Veale and colleagues — writing to help clinicians assess risk in OCD — state it plainly: “there are no recorded cases of a person with OCD carrying out their obsession,” and a person with OCD “is no more likely to act on their intrusions than a person with height phobia is to jump off a tall building.” Their guidance for reassuring patients and families goes further still: “A person with OCD is at no greater risk of causing harm than is any other member of the public (they may even be at a lower risk).”
That paper is also where you find what clinicians actually look at when someone reports a violent thought. These are the features that point toward OCD rather than risk:
| What is assessed | What it looks like in harm OCD |
|---|---|
| Ego-dystonicity | The thought is repugnant to you and clashes with everything you value. It is not attractive, gratifying or entertained. |
| Past behaviour | Nothing in your history is consistent with the thought. |
| Avoidance | You keep away from the trigger — locking away knives, refusing to be alone with the baby, not driving. |
| Frequency | The thoughts are constant and intrusive rather than occasional and cued. |
| Distress | The dominant emotion is fear, guilt and horror — not pleasure or anticipation. |
| Motivation to seek help | You want someone to make it stop. You are reading this page. |
The last row is doing more work than it looks. Being terrified of your own thoughts, and going looking for help because of it, is itself part of the pattern that distinguishes an obsession from an intention.
And here is the boundary, stated as plainly as the rest. If a thought of harming someone does not feel like dread but like something you want — if you find yourself planning it, rehearsing it with any sense of satisfaction, or moving toward it rather than away — that is a different situation and it is urgent. It is not a reason for shame, and it is not something to sit with alone. Contact a crisis line (988 in the US, Samaritans on 116 123 in the UK and Ireland, or your local emergency number) or go to an emergency department. The same applies if you are thinking about ending your life.
Harm to others, and harm to self
These get bundled together online and they should not be. Self-harm OCD is its own fear: the intrusive thought that you might jump from a height, step into traffic, or take your own life — when you do not want to, and the thought terrifies you. Veale and colleagues treat it the same way as any other harm obsession: where OCD is clearly established and the thoughts about self-harm or suicide are ego-dystonic, the case should be managed as for any other intrusion about violence. Most people with OCD, they note, would not want to die by suicide precisely because of the harm it would cause others.
Two things sit alongside that, and this is where a purely reassuring article would be doing you a disservice.
The first is that OCD carries genuinely elevated suicide risk in its own right — not because obsessions get acted out, but because living with a severe, poorly-treated, deeply shameful illness is hard. A 2015 systematic review and meta-analysis in Clinical Psychology Review, pooling 30 independent comparisons from 48 studies, found a moderate-to-high association between suicidality and OCD (Hedges’ g = 0.66, 95% CI 0.49–0.82), holding across both suicidal thoughts and attempts. The authors graded the quality of the underlying studies as low, so treat the size as approximate — but the direction is not in doubt.
The second is more specific. In that 2026 meta-analysis, people with OCD who had suicidal ideation were 1.98 times (95% CI 1.47–2.66) more likely to report aggressive obsessions than those without. That is a co-occurrence, not a causal chain: it says nothing about risk to other people. What it does say is that the two travel together often enough that the authors’ own conclusion was a call for “early identification and increased monitoring among individuals experiencing aggressive obsessions.”
Translated out of research language: harm obsessions are not a reason to be feared, and they are not a reason to be waved off either. They are a reason to be properly assessed by someone who knows OCD.
Why it feels so convincing — and why reassurance makes it worse
Harm OCD is exhausting because the compulsions are mostly invisible. There is no hand-washing to point at. Instead there is mental reviewing (did I do something?), checking your own reactions (did that feel like enjoyment?), avoiding knives and heights and being alone with people you love, confessing to a partner, and searching for one more article that says you are safe.
Every one of those is a compulsion, and each one works — for about ten minutes. Then the doubt returns slightly stronger, because you have just taught yourself that the question needed answering. Veale and colleagues put the mechanism in one line: “Rationalising and reassuring are covert compulsions which maintain their obsessions.”
It is also why well-meant help sometimes backfires. The same review documents a case in which a man with OCD sought help for a fear that he might be a paedophile; a concerned clinician arranged a full risk assessment involving social services, and the effect was that “his OCD became worse because his fears that he might be a danger to the child were now apparently confirmed. His doubts increased and he felt unable to seek further help.” The UK’s National Institute for Health and Care Excellence names the same trap in its OCD guideline: intrusive sexual, aggressive or death-related thoughts “are common in people with OCD at any age, and are often misinterpreted as indicating risk.” This is an argument for seeing someone who knows OCD — not for staying quiet.
If you take one practical thing from this article, make it this: the goal is not to establish that you are safe. You cannot get there, because certainty is exactly what the disorder consumes. The goal is to become able to have the thought without needing to resolve it. That is a learnable skill, and it has a name.
What treatment actually looks like
The treatment for OCD, including harm OCD, is Exposure and Response Prevention (ERP) — a form of cognitive behavioural therapy delivered by a trained clinician. In ERP you approach what triggers the obsession, deliberately and gradually, while not performing the compulsion: no checking, no mental reviewing, no reassurance-seeking. The anxiety rises, and then — unaided — it falls. Repeated enough times, the alarm stops firing.
When the compulsions are mental rather than visible, as they usually are in harm OCD, that is explicitly the recommended approach. NICE guideline CG31 recommends that for adults with obsessive thoughts and no overt compulsions, treatment should include “exposure to obsessive thoughts and response prevention of mental rituals and neutralising strategies” (recommendation 1.5.2.2). NICE also scales the intensity: up to 10 therapist hours where impairment is mild, more than 10 where it is moderate, and combined CBT plus medication where it is severe. SSRIs are the recommended first medication for adults with OCD — that is a conversation for a prescribing clinician, not an article, and the choice and dose belong with them.
How well does ERP work? Honestly, and with the caveats intact. A 2021 systematic review and meta-analysis in Comprehensive Psychiatry pooled 36 randomised controlled trials covering 2,020 patients and found a large effect in favour of CBT with ERP against control conditions (Hedges’ g = 0.74, 95% CI 0.51–0.97). It was clearly better than psychological placebo. It was not better than other active psychological therapies in the eight trials making that comparison, and the authors flagged a real problem: only 8 of the 36 studies were at low risk of bias, and effects were much larger where researchers had an allegiance to the treatment. The International OCD Foundation, which calls ERP the first-line psychological treatment, reports an average 60% reduction in OCD symptoms.
The fair summary: ERP is the best-supported thing we have, it genuinely works for a great many people, and the literature behind it is less pristine than the marketing around it. The IOCDF maintains a resource directory of OCD therapists and clinics, including teletherapy, if you want to find someone trained in it.
What self-directed work can and cannot do
Plenty of people reading this are not going to have an ERP therapist next week — because of cost, waitlists, or where they live. So it is worth being exact about what unguided self-help does.
A 2022 systematic review and meta-analysis in the Journal of Medical Internet Research pooled 11 randomised trials (983 participants) of computer-assisted self-help for OCD with no human contact at all. It was more effective than a waiting list or psychological placebo (standardised mean difference −0.47, 95% CI −0.73 to −0.22). It also lost people at nearly twice the rate (dropout risk ratio 1.98, 95% CI 1.21–3.23). And the authors rated the quality of that evidence as very low, citing high risk of bias and inconsistent results. One useful signal did emerge: programmes built around ERP, and lasting more than four weeks, worked better without becoming harder to stick with.
So: a real effect, on weak evidence, with a retention problem. That is enough to make self-directed work worth doing and nowhere near enough to make it a substitute for treatment. Used well, it is the thing that helps you hold the line between appointments — noticing an intrusive thought as a symptom instead of a verdict, and sitting with the discomfort at 3am instead of checking one more time. General coping skills for anxiety help with the distress; they do not replace response prevention, which is the part that actually changes OCD. That is also the honest scope of a tool like aidx.ai, our AI coaching and therapy service: useful company for the practice, not a clinician and not an ERP protocol. We have written separately about where an AI app genuinely fits with OCD, and about when to choose coaching versus therapy.
One warning specific to this territory, and it applies to apps, forums and articles alike: if you are using any of them to get reassured, you are performing a compulsion with a new interface. Notice the difference between learning something and checking again.
When this is not OCD
There is a version of this article that ends with “you are safe, don’t worry.” It would be more comforting and less useful.
The distinction that matters is between thoughts you do not want and thoughts you do. Obsessions are ego-dystonic: unwanted, distressing, resisted. Urges that are ego-syntonic — wanted, entertained, planned, or accompanied by anything other than dread — are a different clinical picture and need different, urgent help. You cannot make that distinction reliably by reading a page about yourself at two in the morning, and you should not try to; the attempt is itself a compulsion. What a page can do is tell you what clinicians look at. Then a clinician looks.
Get assessed if the thoughts are taking real time out of your life, if you are avoiding people or places to manage them, or if you are simply unsure. Go now, rather than after another month of research, if you have thoughts of suicide or self-harm, if the thoughts feel wanted rather than intrusive, or if you have begun to plan anything. In the US: call or text 988. In the UK and Ireland: Samaritans, 116 123. Elsewhere: your local emergency number.
Common questions
Do people with harm OCD act on their thoughts?
The clinical literature says no. Veale and colleagues, writing in Advances in Psychiatric Treatment in 2009, state that there are no recorded cases of a person with OCD carrying out their obsession, and that the collective experience of OCD specialists internationally is that obsessions about causing harm do not lead people to harm others.
Are people with harm OCD dangerous?
No — and the same review puts it more strongly than that: a person with OCD is at no greater risk of causing harm than any other member of the public, and may be at lower risk. The avoidance, the vigilance and the horror that define harm OCD all point away from action, not toward it.
How can you tell harm OCD from being genuinely dangerous?
Clinicians look at the whole pattern, not one feature: whether the thought is repugnant or wanted, whether past behaviour is consistent with it, whether you avoid triggers, how frequent and distressing the thoughts are, and how motivated you are to get help. Fear, avoidance and help-seeking point to OCD. Wanting, planning or gratification point somewhere else and need urgent assessment.
Why does harm OCD feel like I want it?
Because you are monitoring for it. Search your own mind hard enough for a trace of intent and you will find ambiguous material — a flicker of feeling you cannot immediately categorise — and OCD reads ambiguity as confirmation. Rachman’s account is that the disorder lives in the misinterpretation of the thought, not the thought itself. The sensation of “wanting” is usually the checking, not the wanting.
What is self-harm OCD?
The intrusive fear that you might harm or kill yourself when you have no wish to — the pull at the edge of a platform, the thought of the knife in your own hand. Clinically it is handled like any other harm obsession when OCD is established and the thoughts are ego-dystonic. It is distinct from suicidal ideation, which is a wish or intention to die and needs immediate help.
How do I know if I have harm OCD?
You get assessed. This article can describe the pattern; it cannot diagnose you, and no online test can either. A GP or primary-care clinician can refer you, and the IOCDF directory lists clinicians trained specifically in OCD. Trying to reach certainty by yourself is the disorder’s favourite trap.
Can harm OCD be cured?
“Cure” is the wrong frame, and chasing it tends to feed the problem. OCD is treatable, and treatment works well for many people. What that looks like in practice is that intrusive thoughts stop being emergencies — they still turn up occasionally, and they stop mattering.
What medication is used for harm OCD?
NICE recommends an SSRI as the initial pharmacological treatment for adults with OCD, alone or combined with CBT depending on severity. Which one, at what dose, and for how long is a decision for a doctor who can assess you.
How do I stop the thoughts?
You don’t, and that is the point. Trying to suppress, neutralise or argue down an intrusive thought is the compulsion that keeps it in place. ERP works in the opposite direction: let the thought be there, don’t answer it, and let the alarm burn out on its own.
Last reviewed: September 2026
References
- Fawcett, E. J., Morris, Q., Lahey, C., Corran, C., Krause, S., Bishop, O. C., Rash, J. A., Carter, J., & Fawcett, J. M. (2026). The prevalence and predictors of aggressive obsessions in obsessive-compulsive disorder: A meta-analytic review. Journal of Psychiatric Research, 195, 264–283.
- Veale, D., Freeston, M., Krebs, G., Heyman, I., & Salkovskis, P. (2009). Risk assessment and management in obsessive–compulsive disorder. Advances in Psychiatric Treatment, 15(5), 332–343.
- Radomsky, A. S., Alcolado, G. M., Abramowitz, J. S., Alonso, P., Belloch, A., Bouvard, M., Clark, D. A., Coles, M. E., Doron, G., Fernández-Álvarez, H., Garcia-Soriano, G., Ghisi, M., Gomez, B., Inozu, M., Moulding, R., Shams, G., Sica, C., Simos, G., & Wong, W. (2014). Part 1—You can run but you can’t hide: Intrusive thoughts on six continents. Journal of Obsessive-Compulsive and Related Disorders, 3(3), 269–279.
- Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35(9), 793–802.
- Angelakis, I., Gooding, P., Tarrier, N., & Panagioti, M. (2015). Suicidality in obsessive compulsive disorder (OCD): A systematic review and meta-analysis. Clinical Psychology Review, 39, 1–15.
- Reid, J. E., Laws, K. R., Drummond, L., Vismara, M., Grancini, B., Mpavaenda, D., & Fineberg, N. A. (2021). 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, 106, 152223.
- Imai, H., Tajika, A., Narita, H., Yoshinaga, N., Kimura, K., Nakamura, H., Takeshima, N., Hayasaka, Y., Ogawa, Y., & Furukawa, T. (2022). Unguided computer-assisted self-help interventions without human contact in patients with obsessive-compulsive disorder: Systematic review and meta-analysis. Journal of Medical Internet Research, 24(4), e35940.
- National Institute for Health and Care Excellence (2005). Obsessive-compulsive disorder and body dysmorphic disorder: treatment. Clinical guideline CG31.
- International OCD Foundation. Exposure and Response Prevention (ERP), and the IOCDF Resource Directory.
This article is general information about obsessive-compulsive disorder and is not medical advice or a substitute for care from a qualified professional. OCD is treatable; if you recognise yourself here, speak to a doctor or a clinician trained in Exposure and Response Prevention about an assessment. If you are in crisis, having thoughts of suicide or self-harm, or feel at risk of acting on a thought of harming someone, contact emergency services or a crisis line immediately — call or text 988 in the US, call Samaritans on 116 123 in the UK and Ireland, or use your local emergency number.
Related reading: Relationship OCD (ROCD): When the Doubt Is the Symptom, Not the Answer
