Short answer: yes, on average therapy helps clearly more than waiting or usual care. In depression trials, about 41 percent of people saw their symptoms at least halved after roughly two months, against about 17 percent with usual care. So it helps many people, but not everyone, and not always the first time.
What the numbers say
Depression is the best-studied problem, with more than 1,100 randomised trials of psychotherapy. A 2021 analysis looked at outcomes people actually recognise, about two months after the start:
- Response (symptoms at least halved): 41 percent with therapy, 17 percent with usual care, 16 percent on a waiting list.
- Remission (symptoms largely gone): about one in three with therapy, 7 to 13 percent in the comparison groups.
- Getting worse: 5 percent with therapy, 12 to 13 percent in the comparison groups.
In plain terms: out of 100 people, about 24 more improve clearly with therapy than with usual care. And some people improve without therapy, so not all of the 41 can be credited to it. Across 427 trials, the average effect was 0.69, which researchers read as between medium and large.
Does therapy work for anxiety, PTSD or OCD?
A 2024 analysis of 441 trials and 33,881 patients gives the share of people whose symptoms at least halved:
- Depression: 42 percent
- PTSD: 38 percent
- OCD: 38 percent
- Panic disorder: 38 percent
- Generalised anxiety: 36 percent
- Social anxiety: 32 percent
- Specific phobia: 32 percent
- Borderline personality disorder: 24 percent
Therapy beat the comparison groups for all of these except borderline personality disorder, where the difference was not statistically significant in this analysis. The authors call the results effective, and the absolute rates modest. Both are true.
Why the honest number is smaller than the headlines
- Waiting lists flatter therapy. In a 2025 series on cognitive behavioural therapy (375 trials), effects against a waiting list were always above 0.94. Against usual care they ranged from 0.22 to 1.13.
- A placebo pill is a tougher test. Against a placebo pill with regular contact (10 trials, 1,240 patients), the effect for depression was 0.25: about one extra person benefits for every seven treated. The authors note that this is comparable to antidepressants.
- Many studies are weak. In that same 2025 series, only about 10 percent of the trials had a low risk of bias, and 56 percent a high risk.
None of this means therapy does not work. It means a realistic expectation is better than a promise.
Which type of therapy works best?
For depression, the main types work about equally well: a 2021 comparison of 331 trials and 34,285 patients found few significant differences between them. On average, group therapy and guided self-help were not less effective than individual therapy, and more than 10 sessions was not more effective than fewer.
The relationship matters too. Across 295 studies, a better working relationship with the therapist went together with better results (correlation 0.28). That is a link, not proof of cause, because feeling better can also make the relationship feel better. Still, it is a good reason to look for someone you can be honest with. Not sure who that is? Read about the difference between coaching and therapy, or the myths that keep people from going.
Therapy or medication?
For depression, therapy and medication work about equally well in the short term. After a year, therapy does better, and the combination beats either one alone. Are you taking medication? Do not stop on your own, but discuss it with your doctor. A psychologist or a psychiatrist can each play a different part.
Does therapy work for men?
There is no sign that it works less well. In 14 trials with 1,766 people with depression, men and women did not respond differently to cognitive behavioural therapy compared with medication, and the average differences between them were under one point on a common depression scale. That study did not cover every type of therapy. The bigger barrier may be starting: read more about men and mental health.
Can therapy make things worse?
It happens, but in the depression trials it happened less often with therapy (5 percent) than without (12 to 13 percent). Therapy can feel heavier for a while, and about one in five people stop early. If it keeps getting worse, or you do not feel safe with your therapist, say so or look for someone else.
How to give therapy a fair chance
No study can tell you in advance what will work for you. This is practical advice, not a research finding.
1. Agree on a goal
What should be different in a few months? Say it out loud in one of the first sessions.
2. Check in together
After a few sessions, ask: what has changed, and what are we doing next?
3. Say it when it does not click
A good therapist can hear this. It often makes the work better.
4. Changing is allowed
A different person or approach is not a failure. Many people do not respond to their first attempt.
Not sure what kind of help fits? The which help test gives a first direction. In danger right now? Call US: 988 | UK: 116 123 | EU: 112.
Talk to someone who has space now
Tell us in a few sentences what is going on and get free suggestions of screened psychologists and counsellors who have space now, in your own language, near you or online. No referral needed.
Frequently asked questions
On average, yes. In depression trials, about 41 percent of people saw their symptoms at least halved after roughly two months, against about 17 percent with usual care. It does not work for everyone, and absolute results are modest.
Trials often measure after about two months. For depression, more than 10 sessions was on average not more effective than fewer. That is an average, so discuss with your therapist what is realistic for you.
For depression they work about equally well in the short term. After a year therapy does better, and the combination beats either alone. Do not stop medication without talking to your doctor.
You are not alone, because many people do not respond to a first attempt. Talk about it with your therapist, consider another person or approach, and ask your doctor about other options.
Sources
- Cuijpers et al. (2021), Acta Psychiatrica Scandinavica: response, remission and deterioration in depression.
- Cuijpers et al. (2024), World Psychiatry: response rates for eight mental disorders.
- Cuijpers et al. (2025), JAMA Psychiatry: cognitive behavioural therapy across disorders, control groups and risk of bias.
- Cuijpers (2025), Tidsskrift for Norsk psykologforening: overview of the depression trials, session number, format and comparison with medication.
- Cuijpers et al. (2014), Psychological Medicine: psychotherapy against a placebo pill.
- Cuijpers et al. (2021), World Psychiatry: comparison of the main types of therapy for depression.
- Flückiger et al. (2018), Psychotherapy: the working relationship and results, 295 studies.
- Cuijpers et al. (2014), Depression and Anxiety: men and women in depression treatment.
- Swift & Greenberg (2012), Journal of Consulting and Clinical Psychology: dropping out of therapy.

