Objective: No recent meta-analysis has examined the effects of cognitive-behavioural therapy (CBT) for adult depression. We decided to conduct such an updated meta-analysis. Methods: Studies were identified through systematic searches in bibliographical databases (PubMed, PsycINFO, Embase, and the Cochrane library). We included studies examining the effects of CBT, compared with control groups, other psychotherapies, and pharmacotherapy. Results: A total of 115 studies met inclusion criteria. The mean effect size (ES) of 94 comparisons from 75 studies of CBT and control groups was Hedges g = 0.71 (95% CI 0.62 to 0.79), which corresponds with a number needed to treat of 2.6. However, this may be an overestimation of the true ES as we found strong indications for publication bias (ES after adjustment for bias was g = 0.53), and because the ES of higher-quality studies was significantly lower (g = 0.53) than for lower-quality studies (g = 0.90). The difference between high- and low-quality studies remained significant after adjustment for other study characteristics in a multivariate meta-regression analysis. We did not find any indication that CBT was more or less effective than other psychotherapies or pharmacotherapy. Combined treatment was significantly more effective than pharmacotherapy alone (g = 0.49). Conclusions: There is no doubt that CBT is an effective treatment for adult depression, although the effects may have been overestimated until now. CBT is also the most studied psychotherapy for depression, and thus has the greatest weight of evidence. However, other treatments approach its overall efficacy.