We review the literature on the effectiveness of the monoamine oxidase inhibitors (MAOIs) and present metaanalyses of controlled trials comparing the FDA-approved MAOIs with both placebo and comparator tricyclic antidepressants. For outpatients, metaanalyses with intent-to-treat samples revealed generally comparable overall efficacy for phenelzine, isocarboxazid, and tranylcypromine. Drug-placebo differences were 29.5% (+/- 11.1%) (phenelzine; nine studies), 41.3% (+/- 18.0%) (isocarboxazid; three studies), and 22.1% (+/- 25.4%) (tranylcypromine; three studies). For inpatients, phenelzine was 22.3% (+/- 30.7%) (five studies) more effective than placebo, whereas the isocarboxazid-placebo difference was lower (15.3%) (+/- 12.6%). Both phenelzine and isocarboxazid were significantly less effective than comparator tricyclics for inpatients, whereas tranylcypromine has not been adequately studied. Both phenelzine and tranylcypromine appear to be more effective than tricyclics in depressed outpatients with atypical features. Monoamine oxidase inhibitors are also effective treatments for outpatients who have failed to respond to tricyclic antidepressants. Our review also suggests (1) the FDA-approved MAOIs treat a somewhat different group of patients than tricyclics; (2) more severely depressed inpatients may not respond as well to MAOIs as to tricyclics; and (3) because of preferential MAOI responsivity, atypical or anergic depressions may be biologically different than classical depressions.