Cognitive Behavioral Therapy for Depression: What Trials Show

Cognitive behavioral therapy for depression is a structured, time-limited, present-focused psychotherapy that helps people change unhelpful thinking patterns and the behaviors that keep low mood going^[4]. It is by far the most studied psychological treatment for depression: a 2023 meta-analysis in World Psychiatry pooled 409 randomized trials with 52,702 patients, the largest meta-analysis yet of a single psychotherapy for a mental disorder^[1].

NIMH names CBT and interpersonal therapy as evidence-based therapies for depression, and the American Psychological Association's 2019 guideline recommends CBT for adults^[2][3]. How common depression and major depressive disorder are, by state and population, is tracked on our pillar pages; this guide covers how CBT works and what the trials found.

CBT was originally developed to treat depression, in work by Aaron T. Beck published in 1967 and 1979, and was later adapted for anxiety, substance use, eating disorders and other conditions^[4]. The VA's therapist manual describes a typical course as about 16 sessions held weekly or every other week, and notes that many patients improve substantially after 4 to 18 sessions^[4].

The model behind it holds that emotions are shaped by thoughts and behaviors, and that the two feed each other in depression. Continued inactivity can strengthen beliefs such as being incompetent or that life is meaningless, which leads to further inactivity^[4].

Across 409 trials, CBT for depression had a moderate to large effect compared with control conditions such as care as usual and waitlists (g = 0.79), and the difference was still significant 6 to 12 months later^[1]. In practical terms, 42% of patients receiving CBT responded to treatment versus 19% in control groups, a number needed to treat of 4.7^[1].

Remission rates were 36% with CBT and 15% in control conditions, a number needed to treat of 3.6^[1]. Response rates differed significantly across types of control condition and were lowest for waitlists^[1].

The estimate shrinks under stricter tests. Restricted to trials at low risk of bias, the effect was g = 0.60, and after adjustment for publication bias it was g = 0.47^[1]. Heterogeneity was high in most analyses, meaning results varied widely from trial to trial^[1].

The effect has not faded over time: the authors found no decline in CBT's effect size when comparing trials published before 2001, from 2001 to 2010, and after 2011^[1].

Sessions follow a set structure in the VA protocol: a brief mood check, agenda setting, a review of the previous session's homework, discussion of agenda items, and a new homework assignment^[4]. The protocol pairs these techniques with a case conceptualization, an individualized map of the beliefs, situations and behaviors driving each patient's symptoms^[4].

The behavioral strategies include activity monitoring, activity scheduling, behavioral activation and graded task assignment^[4]. They target what behavioral theory associates with depression: a low rate of rewarding, reinforcing activity^[4].

The cognitive strategies work on automatic thoughts, the thoughts that arise quickly in response to a situation. Patients learn to identify these thoughts, weigh them on three-column and five-column thought records, and use coping cards and a shorthand the manual calls “catch it, check it, change it”^[4]. The manual's appendix supplies the patient worksheets and forms used for these exercises^[4].

Later work moves to core beliefs and problem-solving, then to reviewing skills, tapering sessions and ending treatment^[4]. NIMH notes that newer versions of CBT add mindfulness principles or target specific symptoms such as insomnia^[2].

At the end of treatment, CBT and antidepressant medication perform about the same: the 2023 meta-analysis found no significant difference between them in the short term^[1]. At 6 to 12 months, CBT was significantly more effective (g = 0.34), although that comparison rested on few trials and was not significant in every sensitivity analysis^[1].

Combining CBT with medication beat medication alone in the short term (g = 0.51) and the longer term (g = 0.32), but the combination was not more effective than CBT alone at either point^[1]. How many Americans take antidepressants is covered separately.

Severity shapes how treatment is usually sequenced. NIMH says psychotherapy is often tried first for milder depression, while people with moderate or severe depression usually are prescribed medication as part of the initial plan^[2].

Trial evidence in severe depression is closer than that sequence suggests. In a 2005 trial of 240 outpatients with moderate to severe major depressive disorder, 58% responded to cognitive therapy and 58% to medication at 16 weeks, with remission at 40% and 46%^[5]. The authors cautioned that this result may depend on a high level of therapist experience^[5].

Relapse after treatment ends

The same research group then followed responders after acute treatment. Patients withdrawn from cognitive therapy relapsed at a rate of 30.8% over the next 12 months, compared with 76.2% of medication responders switched to placebo and 47.2% of those who kept taking medication^[6].

Mindfulness-based cognitive therapy (MBCT), a CBT offshoot built for relapse prevention, shows a similar pattern. A 2016 individual patient data meta-analysis of 9 trials and 1,258 patients with recurrent depression found MBCT lowered relapse risk over 60 weeks compared with no MBCT (hazard ratio 0.69) and compared with active treatments including maintenance antidepressants (hazard ratio 0.79)^[7].

Depression therapy can mean psychotherapy, medication or both: NIMH says treatment typically involves psychotherapy delivered in person or virtually, medication, or both^[2]. Antidepressants usually take 4 to 8 weeks to work^[2].

CBT is one of several psychotherapies for depression with guideline support. The APA's 2019 guideline recommends seven psychotherapy interventions, including behavioral therapy, cognitive therapy, CBT, interpersonal psychotherapy, psychodynamic therapy and supportive therapy, along with second-generation antidepressants (SSRIs, SNRIs or NDRIs)^[3].

Interpersonal therapy (IPT) focuses on relationships and life events that affect mood, and NIMH lists it beside CBT as an evidence-based therapy for depression^[2]. NIMH adds that psychodynamic therapy used for a limited time may also help some people^[2].

CBT's lead over these alternatives is small: in 87 head-to-head studies, CBT beat other psychotherapies by g = 0.06, a difference that became non-significant in most sensitivity analyses^[1]. The authors concluded that CBT's superiority over other psychotherapies for depression does not emerge clearly^[1]. By this measure, the case for CBT rests more on the depth of its evidence than on superior results.

For depression that does not improve after at least two antidepressants, NIMH describes options such as esketamine nasal spray and adding a second medication class^[2]. Brain stimulation is covered in our ECT guide.

Self-guided CBT works, with smaller effects. Unguided self-help CBT, delivered through books or internet programs without therapist support, had a small to moderate effect on depression (g = 0.45) across 36 studies with 11,720 participants^[1].

Most trials test CBT delivered individually, in groups or as guided self-help, and the authors' earlier work found no significant differences between those formats^[1]. NIMH states that psychotherapy can be effective in person or via telehealth, and that providers may supplement therapy with apps or other digital tools^[2].

Adoption and engagement data for these tools are in our mental health apps statistics.

CBT also works as an add-on for treatment-resistant depression. In the UK CoBalT trial, 469 primary care patients whose depression had not responded to at least 6 weeks of antidepressants were randomized to usual care or CBT plus usual care^[8]. At 6 months, 46% of the CBT group responded versus 22% with usual care alone^[8].

Response in CoBalT meant at least a 50% reduction in Beck Depression Inventory scores^[8]. Its entry criterion was looser than NIMH’s definition of treatment-resistant depression, which requires no improvement after at least two antidepressants^[2].

Results are smaller or less certain in some groups. The effect in children and adolescents (g = 0.41) is about half the overall estimate, and the estimate for inpatient and nursing home settings (g = 0.65) rests on only 10 studies^[1]. Youth trends are tracked in our teen depression statistics.

Therapist skill may matter. In the 2005 moderate-to-severe trial, medication outperformed cognitive therapy at one of the two sites, a gap the authors linked partly to differences in therapist experience^[5].

In 2024, 14.0% of U.S. adults received counseling or therapy from a mental health professional in the past 12 months, while 19.3% took medication for their mental health, according to the National Health Interview Survey^[9]. Women (16.9%) were more likely than men (10.9%) to receive counseling or therapy^[9].

Use falls sharply with age: 20.2% of adults aged 18 to 29 received counseling or therapy, compared with 7.5% of those aged 65 to 74 and 5.0% of those 75 and older^[9].

All Marketplace health plans must cover behavioral health treatment, such as psychotherapy and counseling, as an essential health benefit, and cannot deny coverage or charge more for a pre-existing mental health condition^[10]. Parity protections mean limits on mental health care, including copayments, visit limits and prior authorization, generally can’t be more restrictive than limits on medical and surgical care^[10].

Coverage rules do not settle out-of-pocket cost or wait times. Those are tracked in our cost of therapy statistics and therapist shortage data.