The cognitive model of depression suggests that cognitive therapy (CT) improves major depressive disorder (MDD) in part by changing depressive cognitive content (e.g. dysfunctional attitudes, hopelessness). The current analyses clarified: (1) the durability of improvements in cognitive content made by acute-phase CT responders; (2) whether continuation-phase CT (C-CT) or fluoxetine (FLX) further improves cognitive content; and (3) the extent to which cognitive content mediates continuation treatments’ effects on depressive symptoms and major depressive relapse/recurrence.

Out-patients with recurrent MDD who responded to acute-phase CT (n = 241) were randomized to 8 months of C-CT, FLX or pill placebo (PBO) and followed for an 24 additional months. Cognitive content was assessed approximately every 4 months using five standard patient-report measures.

Large improvements in cognitive content made during acute-phase CT were maintained for 32 months, with 78–90% of patients scoring in normal ranges, on average. Cognitive content varied little between C-CT, FLX and PBO arms, overall. Small, transient improvements in cognitive content in C-CT or FLX compared with PBO patients did not clearly mediate the treatments’ effects on depressive symptoms or on major depressive relapse/recurrence.

Outpatients with recurrent MDD who respond to acute-phase CT show durable improvements in cognitive content. C-CT or FLX may not continue to improve patient-reported cognitive content substantively, and thus may treat recurrent MDD by other paths.