Background: This study aimed to identify subgroups for whom supported self-help preventive cognitive therapy (S-PCT) is more (cost)effective than treatment as usual (TAU) in preventing relapse and recurrence of major depression. Methods: We conducted a randomized controlled trial in which 248 remitted, recurrently depressed participants were randomized to S-PCT (n = 124) or TAU (n = 124). Clinical outcome was relapse or recurrence of major depressive disorder (SCID-I). We tested the moderating effects on relapse or recurrence of age, gender, education level, residual depressive symptoms, number of previous episodes, age of onset, antidepressant medication, somatization, and self-efficacy with logistic regression analyses adjusted for baseline values of depressive symptoms. We examined moderating effects on costs using linear regression analyses adjusted for baseline costs. A stratified cost-effectiveness analysis was performed to tease out differences in cost-effectiveness between subgroups. Results: We found no moderating effect on relapse or recurrence for any of the potential moderators. For costs, the number of previous depressive episodes was identified as a moderator. At a willingness-to-pay of 16,000€, the probability that S-PCT was cost-effective compared to TAU was 95% for participants with 2–3 episodes and 11% for participants with ≥4 episodes. Limitations: Participants and counselors were not blinded. The study was primarily designed to assess the (cost)effectiveness of S-PCT and not to conduct moderation analyses. Conclusions: S-PCT was effective in preventing relapse or recurrence of depressive disorders in a broad range of participants, but is more likely to be cost-effective in participants with 2–3 episodes than ≥4 episodes. This indicates that S-PCT can best be offered to participants with fewer previous depressive episodes.