Treatment protocols compared

Five well-studied treatments for depression, each shown week by week as it was delivered in its main trial. Same lanes, same time axis, so you can see how they differ in what you do, how long it takes, and how much of the work happens with a therapist.

Under clinical review. Written from the sources listed below; updated November 7, 2026.

These are the versions tested in research. The therapist you see may run things differently, and that's normal. Use this to understand what each approach asks of you and what the evidence says, not as a script.

Click a bar or marker for details.

Lane

Sessions and contact

Learning how it works

Assessment and planning

Doing: activity and behavior

Thinking: thoughts and beliefs

Relationships

Mindfulness practice

Practice between sessions

Staying well afterwards

Follow-up in the trial

Progress measured

Click a bar, a session tick or a progress marker to see what it is and where the detail comes from.

With a therapistOn your ownTherapist at a distanceFollow-up onlyOne sessionClinician-rated measureSelf-report? Detail still to check

Cognitive therapy for depression

Beck's manual, as delivered in DeRubeis et al. 2005 and Hollon et al. 2005

Format: One-to-one with a therapist, 50-minute sessions, face to face
Length: 16 weeks: twice a week for the first 4 weeks, once or twice a week for the middle 8, weekly for the last 4. About 20 to 25 sessions
Therapist time: About 17 to 21 hours, plus up to 3 booster sessions in the following year
How it works: You learn to notice, test and change the negative thoughts and beliefs that drive low mood, and practice it between sessions until it's a habit you keep after therapy ends
Main result: After 16 weeks, 58% improved substantially with cognitive therapy and 58% with medication (240 people; placebo 25% at week 8). In the year after stopping, 31% of the therapy group relapsed, compared with 76% of those taken off medication and 47% of those who kept taking it

What this means for you

Expect a structured hour: you set an agenda, review last week's practice, work on one problem, and agree what to try before next time.
Early sessions focus on getting active again and seeing how thoughts affect mood. Middle sessions teach you to catch and test negative thoughts. Later ones tackle deeper beliefs and plan how to stay well.
It was tested on people with moderate to severe depression, many with long-standing or repeated episodes, aged 18 to 70.
Its main advantage showed up after therapy ended: people who had done cognitive therapy were much less likely to relapse than people who stopped medication, and the benefit lasted into a second year.

Worth knowing

Therapist skill mattered. Therapy matched medication at the site with highly experienced cognitive therapists; at the other site, where two therapists were newer, medication did better on full remission.
About 6 in 10 responded within 16 weeks and about 15% dropped out. Roughly 4 in 10 did not get enough better by the trial's definition.
The relapse comparison was small, about 35 people per group, and the medication group was stopped after 4 months, which no clinician would recommend. It doesn't tell you what happens with therapy plus medication.

Sources

Behavioral activation for depression

Martell's manual, as delivered in Dimidjian et al. 2006 and Dobson et al. 2008; COBRA (Richards et al. 2016) for the UK version

Format: One-to-one with a therapist, face to face. In the UK trial, delivered by junior mental health workers after 5 days of training
Length: 16 weeks, up to 24 sessions of 50 minutes (US trial, about 20 on average). UK trial: up to 20 sessions of 60 minutes plus 4 optional boosters; 11.5 attended on average
Therapist time: About 17 to 20 hours in the US trial; 11.5 hours on average in the UK trial
How it works: Depression is kept going by pulling back from life. Scheduled, goal-directed action, done by plan rather than by mood, brings back the things that lift mood
Main result: US trial (241 people): for more severely depressed people, behavioral activation matched an antidepressant and both beat cognitive therapy; gains held over two years without further sessions. UK trial (440 people): a year later it matched full CBT, with two-thirds recovered in both groups, at 21% lower cost

What this means for you

Expect something practical: you keep a daily record of what you do and how you feel, then plan small, specific activities each week and build up. Sessions review the record and plan the next steps.
It doesn't focus on challenging your thoughts. Rumination is handled by changing what you do when it starts.
It suits people who want a concrete, action-based approach, who have withdrawn from activities, or who can't get or don't want CBT or medication.
Because it can be delivered well by less specialized staff, it's often easier to find and cheaper.

Worth knowing

About a third of people in the UK trial attended fewer than 8 sessions, and 20 to 30% didn't improve meaningfully with either behavioral activation or CBT.
The finding that it beat cognitive therapy for severe depression came from a small subgroup, and a 2019 reanalysis didn't confirm it. The two-year follow-up had only 13 behavioral activation patients in its second year.
Most people in the UK trial (78%) were also taking antidepressants, so its results reflect therapy added to usual medication.

Sources

Interpersonal psychotherapy for depression

Klerman and Weissman's manual, as delivered in the NIMH collaborative trial (Elkin et al. 1989; Shea et al. 1992)

Format: One-to-one with a therapist, weekly, face to face, in three phases
Length: 16 weeks; 16 to 20 sessions of about 50 minutes
Therapist time: About 13 to 17 hours
How it works: Depression is treated by working on one current problem between you and other people: a loss, a dispute, a life change, or isolation. Better relationships and support relieve the symptoms
Main result: In the largest head-to-head trial (250 people), interpersonal therapy matched CBT and came close to the antidepressant imipramine; 43% of people assigned to it met a strict recovery cutoff by week 16. Across 38 trials it beats no treatment or usual care by a moderate to large margin and equals other therapies

What this means for you

Expect about 16 weekly sessions over about 4 months. The first 2 or 3 map your relationships and settle on one focus. Most of the course works on that focus. The last 2 or 3 review gains and plan for the future.
It fits best when the depression is tied to a loss, an ongoing conflict with someone close, a big life change such as a new job, divorce, a new baby, retirement or illness, or long-standing isolation.
There are no thought records or formal homework. Instead you talk through recent interactions, rehearse conversations, and try them out between sessions.

Worth knowing

A 16-week course isn't a cure for most people. In the big trial about a quarter of those assigned to interpersonal therapy were well at the end and stayed well for 18 months, and a third of those who recovered relapsed.
Medication had a small edge on average across trials, and for more severe depression the big trial favored imipramine, with some support for interpersonal therapy.
If your main difficulties are with concentration and work rather than with people, it may fit less well. Recurrent depression may need maintenance sessions.

Sources

Elkin et al. (1989). NIMH Treatment of Depression Collaborative Research Program: general effectiveness of treatments. Arch Gen Psychiatry 46:971
Shea et al. (1992). Course of depressive symptoms over follow-up: NIMH TDCRP. Arch Gen Psychiatry 49:782
Cuijpers et al. (2011). Interpersonal psychotherapy for depression: a meta-analysis. Am J Psychiatry 168:581
Markowitz & Weissman (2004). Interpersonal psychotherapy: principles and applications. World Psychiatry 3:136
Weissman, Markowitz & Klerman (2018). The Guide to Interpersonal Psychotherapy. Oxford University Press

Mindfulness-based cognitive therapy, for staying well

Segal, Williams and Teasdale's program, as delivered in Teasdale et al. 2000, Ma & Teasdale 2004 and the PREVENT trial (Kuyken et al. 2015)

Format: A closed group of about 8 to 15 with one instructor; 8 weekly classes of 2 to 2.25 hours, an individual interview first, an optional full day of practice, and 4 refresher sessions over the next year in many services
Length: 8 weeks of classes, about 16 to 18 hours in the room, plus about 45 minutes of home practice 6 days a week (about 36 hours)
Instructor time: Under 5 hours of instructor time per person, because it's a group
How it works: You learn to notice early mood-linked negative thinking and to relate to thoughts and feelings as passing events rather than facts, so rumination doesn't escalate a low mood into a new episode
Main result: For people with three or more past episodes, relapse over a year fell from 66% to 37% (Teasdale 2000, 145 people) and from 78% to 36% (Ma 2004, 55 people). Against staying on antidepressants, it was about as good: 44% vs 47% relapsed over two years (PREVENT, 424 people). Across 9 trials, 38% relapsed within 60 weeks with MBCT vs 49% without

What this means for you

This is for staying well after you've recovered, not for treating a current episode. The trials enrolled people who were in remission or stable on antidepressants.
It suits people with three or more past episodes. The original trials found benefit only in that group, and none for people with two.
Expect a group, guided meditations such as the body scan, sitting practice and the 3-minute breathing space, an optional day of practice, and a real commitment to daily home practice.
In PREVENT, about 7 in 10 people who attended at least four classes were able to stop their antidepressant, with their doctor managing the taper, and did no worse than people who stayed on it.

Worth knowing

Results depend on doing the daily practice. In PREVENT, people attended 6 of 8 classes on average, and home-practice adherence in studies varies.
It's not a substitute for treating a current episode, and tapering antidepressants should only be done with your prescriber.
PREVENT found no cost saving over medication, and it was not better than medication, only about equal.

Sources

Guided internet CBT for depression

The Swedish program tested in Andersson et al. 2005; module content from a later version of the same material (Flygare et al. 2019)

Format: Text-based self-help on a website, 5 modules, with a named therapist giving brief written feedback by email, plus a moderated discussion forum. Done at home
Length: 8 weeks advised, 10 on average. Each module is unlocked after the quiz on the previous one
Therapist time: About 2 hours per person over the whole course. Roughly 10 minutes a week is typical of guided programs
How it works: You learn and practice behavioral activation and cognitive restructuring from structured text, with short written feedback that keeps you on track
Main result: Beck inventory scores fell from 20.5 to 12.2 in the treatment group vs 20.9 to 19.5 in the control group (117 people; a large effect, d = 0.94), and held at 6 months. Across trials, guided programs (d about 0.6) work far better than unguided ones (d about 0.25), and guided internet CBT performs about the same as face-to-face CBT head to head

What this means for you

It suits people with mild to moderate depression who read and write comfortably, want to work at their own pace at home, and can keep going without a weekly appointment.
Expect 5 to 8 text modules over about 8 weeks, with reading, exercises and a short quiz or homework after each. A named therapist replies by secure message within about a day and unlocks the next module.
The therapist time is small, about 2 hours over the whole course, which is why it's cheap and widely available.
Choose a program with a real human guide and a clear timeline. Unguided apps are much weaker.

Worth knowing

Dropout is the weak point. 37% left the 2005 program (vs 18% of controls), people completed under 4 of 5 modules on average, and only about 65% of people finish guided programs in general. "Too demanding" was the main reason given.
It's not designed for severe depression or for people at risk of suicide; the trial excluded them.
Long-term data are encouraging but thin. In the sibling program, 57% had minimal symptoms 3.5 years later, but over half had sought further treatment, and only 58% were followed up.

Sources

Side by side

Treatment	Format	Length	Therapist time	What you do	Thought work	Between sessions	Follow-up in the trial	Evidence
Cognitive therapy	One-to-one, 50 min	16 weeks, about 20 to 25 sessions	About 17 to 21 h	Activity scheduling early, then behavioral experiments	The core: thought records, then core beliefs	Homework every week	12 months with up to 3 boosters, then a year observed	Matched medication at 16 weeks; fewer relapses afterwards
Behavioral activation	One-to-one, 50 to 60 min	16 weeks, up to 20 to 24 sessions	About 12 to 20 h	The core: diary, scheduled activities, replacing avoidance	Little; rumination handled by changing what you do	Daily diary and planned activities	2 years observed (US); 18 months (UK)	Matched CBT at 12 months (UK) and medication for severe depression (US)
Interpersonal therapy	One-to-one, about 50 min	16 weeks, 16 to 20 sessions	About 13 to 17 h	Work on one relationship problem	None by design	No formal homework; try things out	18 months observed	Matched CBT, close to imipramine; d about 0.6 vs no treatment
MBCT	Group of 8 to 15, 2 to 2.25 h classes	8 weeks plus optional day and refreshers	Under 5 h per person	Daily meditation practice; activity plan at the end	Thoughts as passing events (class 6)	About 45 min a day, 6 days a week	1 to 2 years, with refreshers	Relapse 38% vs 49% across 9 trials; about equal to staying on medication
Guided internet CBT	Online text modules with email feedback	8 to 10 weeks, 5 modules	About 2 h per person	Behavioral activation module, sleep module	Cognitive restructuring module	All of it: reading, exercises, quizzes	6 months	Large effect vs control; guided about equal to face-to-face CBT

Which of these fits you?

Answer a few questions and we'll suggest the kind of help that fits where you are now, and where to get it near you. Your answers stay on this device.

Find your step

How to read the marks: a bar's details say whether the timing is stated in the trial paper, comes from the treatment manual with the week estimated, or still needs checking (shown with a ?). Results are from the trials named on each card; other trials of the same treatment can differ.

This page is general information, not medical advice. Several kinds of therapy work about equally well for most people; choosing one you're comfortable with and can stick to matters more than the type. See Therapy for depression and The steps of care.