SummaryA treatment trial comparing cognitive therapy and pharmacotherapy, alone and in combination, in depressed out-patients, indicated that outcome on cognitive variables was similar to outcome on mood and severity measures, pharmacotherapy being less effective than cognitive therapy or the combined treatment in a hospital and a general practice sample. While combined treatment was superior to cognitive therapy in the hospital outpatients, the two treatments were equivalent in the general practice. Significant effects were obtained on measures of views of self, the world and the future, whose validity was demonstrated. The pattern of change through treatment showed the same order of progress for responders, while non-responders to cognitive therapy tended to do worst. The specificity of treatment effect is discussed.