This study tested the use of a problem solving measure (The Problem Solving Scale: Center for Cognitive Therapy, 1988) as a predictor of outcome in patients receiving cognitive behaviour therapy (CBT) in an NHS psychotherapy service. Lower scores at assessment on the Problem Solving Scale (PSS) were predicted to be associated with poorer outcome for all diagnoses (but particularly depression) and higher dropout. The PSS predicted outcome for depressed patients receiving CBT, but not anxious patients or patients with other disorders. Poorer problem solving was not associated with dropout from therapy. Components of the scale that were most closely correlated with global outcome seemed to reflect the behavioural activation and cognitive restructuring skills taught in CBT for depression. Since this is the second study to connect PSS scores with outcome in depression (Scott, Harrington, House, & Ferrier, 1996), the possibility that aspects of learned resourcefulness are relevant to patients' suitability for CBT remains alive, and further research in this area is indicated.