evIdence For cbt For dePreSSIonThere are more RCTs of CBT for depression than there are for any oth- er psychotherapy for any other disorder, making it the most replicated test of psychotherapy in the literature. Thoma and colleagues (2012) found 120 trials of CBT for depression as of 2010, a number which has since grown. This included comparisons of CBT to a variety of control conditions, including a wait list, treatment-as-usual (TAU), other psy- chotherapies, medication, and the combination of CBT and medication. Most of these trials focused on major depressive disorder (MDD) and over half used the Beckian model. In comparison to a wait list control, CBT has shown large effects, with Cohen’s d = 0.90 (Thoma et al., 2012).1 This demonstrates that CBT does better than doing nothing. In compar- ison to TAU, medium effects were found, with d = 0.40. However, this finding is ambiguous, as Wampold et al. (2011) demonstrated that TAU can consist of a wide array of control conditions that range from very inactive to bona fide treatment. In comparison to medication, no differ- ence was found, with an effect size close to zero at d = 0.10. This sug- gests that on average, the effects of CBT for depression are comparable to that of antidepressant medication. In comparison to other forms of psychotherapy, likewise the effect was nonsignificant and close to zero, with d = 0.05. In sub-analyses, in which comparisons between CBT and BT were removed, the results were the same. Alternatively, a compari-son between CBT and CBT with medication, the combined treatment had larger effects, showing additive effects of the two treatments.Taken together, these findings indicate that CBT is decidedly better than nothing but perhaps no better and no worse than medication or other psychotherapies for treating a major depressive episode. Com- bining CBT with medication appears to have beneficial additive effects. Interestingly, Thoma et al. (2012) found that the quality of the method- ology of the RCTs was inversely related to outcome. This is to say, the better the trial, the worse the outcome for CBT. Additionally, there was evidence of publication bias. This means that the aggregated results may be somewhat inflated due to trials with small sample sizes and poor results going unpublished.