The authors conducted a critical review of the literature on interventions to improve provider recognition and management of mental disorders in primary care, searching the MEDLINE database for relevant articles published from 1966 through May 1998 and finding 48 usable controlled studies (27 randomized controlled trials and 21 quasi-experimental studies). Improved diagnosis of mental disorders was reported in 18 of 23 (78%) of the studies examining this outcome and improved treatment in 14 of 20 studies (70%); clinical improvement in psychiatric symptoms or functional status was documented in 4 of 11 and 4 of 8 (36% and 50%, respectively). Considerable study heterogeneity precluded subjecting the literature synthesis to a formal meta-analysis of pooled results; the authors were therefore unable to demonstrate an association between efficacy of an intervention and any specific variables. A variety of interventions and further research may be effective in improving the recognition and management of mental disorders in primary care.