BACKGROUND The clinically relevant outcomes in treating depression are persistent recovery, relapse, and treatment resistance. METHODS 175 outpatients treated with antidepressants for 6 months were assessed for major depression. Those who had recovered were prospectively monitored for one year to study rates of relapse (at least two weeks of major depression). Those who were depressed at 6 months were monitored for rates of recovery (at least 8 weeks of no major depression). RESULTS 94% of the sample was monitored for one year. Of the 123 patients who were not depressed at 6 months 57 (46%) relapsed. Patients who relapsed were more likely to have a history of recurrent depression, to have residual depressive symptoms, to have a less sustained response to initial treatment, to have avoidant personality disorder symptoms, schizotypal personality disorder symptoms, higher harm avoidance (HA) scores and lower self directedness (SD) scores. Of the 38 patients who were depressed at 6 months 13 (34%) recovered. There were no patient characteristics associated with recovery. CONCLUSIONS The findings apply to moderately depressed outpatients. There was no placebo control. CONCLUSIONS Most patients with depression will recover but many become unwell again within a year. Clinically long term monitoring and sustained efforts to treat patients with major depression seem warranted.