Does therapy actually work?
Yes, it does. For depression, about 41% of people who start therapy see a big improvement in their symptoms. That compares to 17% of people getting usual care from their GP (general practitioner) For most people, where getting help starts. A GP can write a Mental Health Treatment Plan, prescribe medication and refer you on, and some are trained to deliver a small number of structured therapy sessions themselves. More in the jargon decoder , and 16% of people waiting on a list (Cuijpers and colleagues, 2023). It doesn’t work for everyone, and a small number of people come out worse. Who you see usually matters more than the exact method they use.
What the research shows
Most of the good research is about depression. In a large overview of 228 trials involving 23,574 people, psychological treatment consistently beat both usual medical care and waiting lists (Cuijpers and colleagues, 2023).
In those studies, “improved a lot” means a person’s symptom score dropped by about half over a short, fixed course of care. Researchers use that mark so they can compare different trials against each other. It isn’t a line that separates people who were helped from those who weren’t, and most people in the trials didn’t reach it.
About a third of the people who had therapy were no longer depressed by the end. Among people who got standard care or sat on a waiting list, it was 9% to 12%.
How many people get worse
Across the research, about 4 in every 100 people in therapy saw their symptoms worsen. Among people on waiting lists or getting usual care from a GP, it was about 11 in every 100.
Seeing symptoms go up during a trial doesn’t mean therapy caused the downturn. Studies track people across months of ordinary life, and those months bring plenty of outside events. In the same way, some people who improved would have felt better on their own. Comparing the therapy group with the control groups is what separates what the therapy did from what life did.
Symptom scores also miss specific types of harm. Sessions can do damage if things move too fast into something painful, or if a therapist isn’t right for you and won’t adjust when you say so. So can advice that makes a tough situation worse.
Tell your therapist directly if something feels wrong in the room. How to tell your therapist something isn’t working covers how that conversation usually goes. If they don’t adapt, you can change therapists or stop.
How many people stop early
About one in five people stop before their therapy is finished, across 669 studies and 83,834 clients (Swift and Greenberg, 2012).
People drop out for all sorts of reasons:
- Sessions feel uncomfortable or unhelpful.
- Practical hurdles like fees, schedules or travel get in the way.
- People start feeling better sooner than planned and decide they’ve had enough.
No therapist is surprised when someone stops.
Does the type of therapy matter?
For most difficulties, no single Approach (modality) The kind of therapy someone uses, such as CBT or psychodynamic therapy. You'll also see it called a modality. More in the jargon decoder consistently beats the others.
The clearest exception is obsessive compulsive disorder. A specific style of cognitive behavioural therapy called Exposure and response prevention (ERP) A kind of CBT for obsessive compulsive disorder (OCD). You gradually face the things that set off your anxiety, then hold off on the checking, washing or asking for reassurance you'd normally do to bring the anxiety down. More in the jargon decoder is the recommended first treatment, and it has stronger evidence for OCD than other kinds of talking therapy (Öst and colleagues, 2015; Olatunji and colleagues, 2013). You gradually face the things that set off your anxiety. Then you hold off on the checking, washing or asking for reassurance you might normally do to reduce your anxiety.
On the other hand, non-directive supportive counselling came out behind other approaches in depression trials. It still helped people more than no treatment, but it produced smaller gains than more structured methods.
In those trials, researchers usually design “non-directive supportive counselling” as a basic comparison setup — a therapist offering warmth and attention while deliberately holding back any active tools. Person-centred therapy as it’s practised outside a trial is its own approach, with its own training behind it.
The therapist matters more than the method
Which therapist you see accounts for about 5% of the differences in how people turn out, with study estimates ranging from 0.2% to 29% (Johns and colleagues, 2019). That’s still more than the gap between different therapy models.
How well you and your therapist work together makes a difference too. A review of 295 studies found that a stronger working relationship went with better results, across many kinds of therapy (Flückiger and colleagues, 2018). Studies like these show a link rather than proving which causes which. Finding someone you trust and can be straight with can matter more than the name of the type of therapy they use.
You can read more about that in why who you see matters more than what they do.
Does it last?
The benefits of therapy are still measurable 6 to 12 months after starting, which is as far as most trials follow anyone. Research thins out after that.
Depression often returns over a lifetime, and some approaches include a section on preventing that — spotting an early dip, and having a plan ready.
What the research can’t tell you
Out of 331 trials in a major overview, only about a third met every mark for a well-run study. Researchers can adjust for the weak setups, and for trials whose disappointing results were never published. Therapy is still clearly ahead of no treatment afterwards, by a smaller margin than the headline figures suggest.
Trials also test a narrower setup than what you’ll encounter in real life:
- Studies use firm rules about who can take part, usually people with one clearly defined problem. Most people arrive with several things at once.
- Methods like Cognitive behavioural therapy (CBT) A structured, practical form of talking therapy. It's built on a simple idea: what you think affects how you feel, and both affect what you do. More in the jargon decoder get studied far more often because they’re easy to standardise and cheap to run.
- A therapy model being under-tested isn’t the same thing as it being tested and not working.
What happens in Australia
About one in ten Australians used at least one Better Access The Medicare program that gives rebates on mental health sessions with psychologists and some other professionals, through a Mental Health Treatment Plan. More in the jargon decoder service in 2021. About one in twenty had at least one session of psychological treatment (Department of Health, 2022).
In routine clinical records, over half of completed treatment courses showed clear improvement in symptoms and day-to-day functioning. That figure can’t be compared with the trial numbers above. Everyday clinical records have no control group and use different measures.
People use 5.4 sessions on average, well under the ten a Mental Health Treatment Plan allows each year. The scheme also reaches people on higher incomes more than people on lower ones, and remote areas least of all. Free and low-cost options sets out what’s available outside the scheme.
What to do with all this
- Give it a go. Therapy offers a real chance of making life easier to handle, and the gap between therapy and no therapy is the most reliable finding on this page.
- Pick someone you can be straight with. The personal fit matters much more than which approach a therapist trained in.
- Speak up when it isn’t working. Tell your therapist if you feel stuck or misunderstood, so you can adjust things together.
- Don’t treat one bad experience as the answer. Finding the wrong therapist or method doesn’t mean therapy can’t work for you.
If things are difficult right now, you don’t need to settle any of this first. Lifeline is on 13 11 14, the Suicide Call Back Service is on 1300 659 467, and Kids Helpline is on 1800 55 1800 for anyone aged 5 to 25. In an emergency, call 000. Crisis and emergency support has the full list.
Where next
How to find a therapist covers where to search in Australia, how to check someone is Registered For psychologists, psychiatrists and occupational therapists, registration with Ahpra is required, and using the title without it is an offence. Counsellors and psychotherapists can register with their national associations, which is voluntary. More in the jargon decoder , and what to do when nobody has an appointment.
What happens in a first session covers what to expect when you walk in, from the paperwork to working out what you want to be different.