Background: Baseline severity has emerged as a key predictor of acute response to treatments for depression. The goals of this individual patient data meta-analysis were to compare the relapse prevention effects of acute phase cognitive behavioral therapy (CBT) vs. acute phase antidepressant medications (ADM) either continued (-c) or discontinued (-d) and determine whether baseline depression severity moderated these effects. Methods: We included all available relevant randomized trials of CBT versus ADM in adult outpatients with depression. Cox proportional hazards models were used to examine whether treatment condition, baseline severity, and additional characteristics were associated with relapse. Results: Using individual participant data from 5 of 10 published trials (N = 341), CBT (HR = 0.38, 95% CI 0.26–0.57) and ADM-c (HR = 0.48, 95% CI 0.29–0.80) were superior to ADM-d in preventing relapse over 12 months but did not differ from each other (HR = 1.26, 95% CI 0.76–2.09). Baseline severity did not moderate these effects. Conclusions: Regardless of a patient’s baseline symptom severity, CBT and ADM-c both prevent depressive relapse substantially better than medication discontinuation. Given the shorter duration of treatment and equivalent longer-term outcomes, treatment with CBT might be considered a first choice for adults with depression.