A double-blind, placebo-controlled study of 229 patients with a Research Diagnostic Criteria diagnosis of major, minor or intermittent depression was used to compare the clinical profiles of venlafaxine and imipramine in general practice. Venlafaxine produced a significant improvement compared to placebo in symptoms of depression and anxiety as rated by the total MADRS and percentage of responders, the CGI improvement, the CGI severity of illness, the BSA psychic anxiety item and the HSCL. On a number of these measures, venlafaxine was also significantly more effective than imipramine. Venlafaxine was significantly superior to both imipramine and placebo for the SARS total score and the items 'social/leisure' and 'extended family.' A similar proportion of patients discontinued treatment in each group, but fewer patients on venlafaxine discontinued treatment because of an unsatisfactory response.