Background: It is widely agreed that chronic depression is difficult to treat, knowledge about optimal treatment approaches is emerging. Method: A multisite randomized controlled trial was conducted comparing the cognitive behavioral analysis system of psychotherapy (CBASP), a psychotherapy model developed specifically to treat chronic depression (n = 67) with care as usual (CAU; evidence-based treatments, n = 72) over a period of 52 weeks, with 23 sessions on average, in 3 outpatient clinics in the Netherlands. In both arms algorithm-based pharmacotherapy was provided. Patients (aged 18-65) met criteria for a DSM-IV diagnosis of major depressive disorder with diagnostic specifiers (chronic, without interepisode recovery) or with co-occurring dysthymic disorder indicating a chronic course. The Inventory for Depressive Symptomatology (IDS) Self-Report was used as the primary outcome measure. Mixed-effects linear regression analysis was used to compare the changes on the IDS scores between CBASP and CAU. The IDS was administered before treatment, and after 8, 16, 32 and 52 weeks. Results: At week 52, patients assigned to CBASP had a greater reduction of depressive symptoms compared to patients assigned to CAU (t = -2.00, p = 0.05). However, CBASP and CAU did not differ from each other on the IDS after 8 weeks (t = 0.49, p = 0.63), 16 weeks (t = -0.03, p = 0.98) and 32 weeks (t = -0.17, p = 0.86) of treatment. Conclusions: This trial shows that CBASP is at least as effective as standard evidence-based treatments for chronic depression. In the long run, CBASP appears to have an added effect.