This is a 2-year, double-blind, placebo-controlled study of 101 patients, evaluating the relative efficacy of intermittent medication (given only when the patient shows early signs of relapse) compared with moderate doses of maintenance medication for stable schizophrenic outpatients. Patients were dropped from the study if they had three prodromal episodes in 1 year or if an episode lasted more than 9 weeks. Fourteen percent of patients given maintenance treatment were dropped from the study compared with 46% of intermittently treated patients. Relapse rates were 16% for patients given maintenance treatment and 30% for intermittently treated patients, a nonsignificant difference. Intermittently treated patients were receiving significantly less medication, but there were no differences found in drug side effects. There appears to be no advantage in using the intermittent approach, but we found that the use of an early intervention strategy reduced the relapse and rehospitalization rates for these patients.