Getting better

Can I do ERP on my own? And are OCD apps any good?

Maybe there's no OCD specialist within a hundred miles of you. Maybe the waitlist is five months, or this is the year therapy doesn't fit the budget. Here's what going it alone can really do, and where its edges are.

Can I do ERP on my own?

Partly, and the answer has edges. In randomized trials, self-guided exposure and response prevention produces real improvements in obsessive-compulsive disorder, smaller on average than therapist-delivered care, and with more people quitting partway. Clinical guidelines treat structured self-help as a legitimate first step for milder OCD, with escalation to a clinician when it isn't enough.

The reasons behind the question are usually practical rather than philosophical. Perhaps there's no OCD specialist within driving distance, or the nearest one has a waitlist measured in months, or money is the problem this year. Those are real constraints, and I think they've earned a straight answer, not a lecture about how therapy is better.

The steps of ERP themselves are learnable, and people do practice them alone. You put yourself in contact with whatever sets off the doubt, and then you don't do the compulsion afterward. That much is no secret. But studied head to head, going fully alone helps less than doing the same work with a clinician, and more people stop partway. And there are four situations, further down this page, where starting alone is the wrong call altogether.

What does the evidence say about self-help ERP?

Across randomized trials of self-guided programs for obsessive-compulsive disorder, improvements are real but moderate, smaller than in therapist formats, and quitting is more common. Adding even limited clinician contact is associated with larger improvements and better follow-through. Remote therapist-delivered treatment, by contrast, performs comparably to face-to-face care.

The widest view we have is a 2026 network meta-analysis pooling 61 randomized trials of cognitive behavioral therapy (CBT) formats for OCD, 3,710 patients in all. Every format in it beat no treatment, from individual therapy all the way down to fully self-guided programs. The fully self-guided programs came in below most of the other formats, with effects the authors still rate as moderate, and their dropout ran higher. The authors' conclusion cuts both ways, and it's worth having verbatim. Unguided self-help "has lower but still moderate effects in reducing OCD symptom severity, and it holds important potential for assisting a larger number of individuals with OCD who face barriers to accessing treatments."

Those pooled numbers have a fragile spot, and it's worth seeing before you build a plan on them. A meta-analysis of the fully unguided trials alone found a moderate effect among the people who finished, and no significant effect once everyone who started was counted, because so many people stopped partway. And 92% of the included trials were rated at high risk of bias, which is a reason to hold all of it loosely. So self-guided work mostly fails by being abandoned, not by being useless, and the useful question is what keeps a person in it.

Guided self-help, meaning a workbook or program plus brief regular contact with a clinician, showed no significant difference from individual therapy in the 61-trial comparison, though the review rated none of those comparisons high-confidence. Note the word guided. A clinician stays involved, even lightly. And fully remote, therapist-delivered ERP over video performs about as well as face-to-face care, so before you conclude you're out of options because the nearest specialist is hours away, it's worth knowing that distance has stopped being the barrier it used to be. The vetting questions work over video too.

None of this is a workaround the guidelines frown on. The UK's national guideline runs milder OCD on exactly this stepped logic, starting with low-intensity ERP built on structured self-help materials, up to 10 therapist hours, and escalating to intensive therapy or medication "because these treatments appear to be comparably efficacious" when the first step isn't enough. Starting alone isn't cheating. Staying alone after it's clearly not enough is the actual mistake.

Are OCD apps any good?

Apps and online programs for obsessive-compulsive disorder are tools, not treatment. The International OCD Foundation (IOCDF) describes online programs and apps as a beneficial resource when therapy isn't an option, particularly for learning about the disorder and for symptom management. What separates a useful tool from a harmful one is whether it teaches you to sit with uncertainty or sells you certainty on demand.

To be fair to the category, the IOCDF's guidance on low-cost options says that when therapy isn't accessible, "online programs and apps can be a highly beneficial resource," designed around symptom management and CBT principles. Their treatment guide lists self-help resources as a complement to care, and nobody serious claims a phone can't help at all.

A good tool teaches you how the cycle works and gives your daily practice some structure, and it can hold your plan between sessions or while you wait for care. What no tool can do is assess you, adapt exposures to your presentation on the fly, or handle what surfaces when the work gets deep. That part belongs to a clinician, and the good tools admit it.

There's a one-question test I'd have you run on any app in this space. What does it do when you're scared? A tool built for OCD teaches you to let the doubt sit unanswered. A tool built for engagement answers you, warmly, every time you ask, at 2 AM, in fresh words each round. OCD is an illness of uncertainty intolerance, so an app handing out certainty on demand is running the reassurance ritual with you, just at a scale you'd never manage alone. An honest tool for this disorder will sometimes feel unsatisfying in the moment. That's the clinical design, not a flaw.

Full disclosure, since you're reading this on our site. Conicia, the program this library belongs to, is one of the tools in this category, and every line above applies to it as much as to anything else. Conicia is a step-by-step program for understanding OCD and building skills for daily life, rooted in principles of evidence-based approaches, built to run on its own or alongside care with a clinician. Conicia does not deliver ERP and is not therapy. If what you need is ERP itself, what you need is a clinician trained in it, and this library will keep telling you that.

When is on-my-own not the right call?

Self-guided work fits milder obsessive-compulsive disorder, by guideline and by evidence. It's the wrong starting point when symptoms are severe or worsening quickly, when serious depression or a crisis is also present, or when you can't yet tell your OCD from a feared emergency. And if working the program has itself turned into a compulsion, that's information worth taking to a clinician.

The limits show up in four places:

• Severity. The stepped-care logic starts self-help at mild functional impairment. If OCD is running your days, deciding what you can touch and where you can go and how long the morning takes, start with a clinician, not a workbook.
• What's riding along. If serious depression is in the picture, or a crisis, or any thoughts of not wanting to be alive, those outrank the OCD plan. 988 or HOME to 741741 come first, then a clinician.
• Uncertainty about what this is. If you can't yet tell whether tonight's fear is OCD or a real emergency, that call is a clinician's to make, in a conversation. No page or app can make it, and this one is no exception.
• The program became the compulsion. Maybe it's re-reading the same lesson for the relief it brings, or restarting the plan so you can do it "right" this time. Maybe it's checking the app the way you used to check the lock. If you recognize yourself there, you haven't failed the tool. You've found exactly the pattern a clinician can work with.

The realistic path for most people reading this page isn't alone forever, and it isn't specialist tomorrow either. It's starting now, with the right kind of materials, while the search for the right clinician runs alongside. Both can begin this week, and neither one has to wait for the other.