An extensive series of investigations over the past 2 decades clearly demonstrate that mental disorders are present in approximately 25% of primary care patients and that physicians underdiagnose these illnesses. The factors producing this bias are poorly understood and should be focused upon in future research. Also requiring much more study is the efficacy of pharmacologic and psychosocial treatments initially validated with psychiatric populations. Clinical trials should determine whether standardized interventions can be utilized with medical patients whose symptom profiles and organic comorbidity may differ from those of psychiatric populations.