Previous studies have reported a therapeutic window (i.e., a curvilinear relationship between clinical response and drug level) for haloperidol concentrations in serum or plasma. The authors treated 30 acutely decompensated schizophrenic inpatients with a fixed dose of haloperidol (.4 mg/kg/day). After 6 weeks there was no statistically significant correlation between clinical improvement and serum haloperidol concentration. Fifteen subjects with serum concentrations of 5-15 ng/ml did not differ in clinical improvement compared with 15 subjects who had concentrations above 15 ng/ml. These data are consistent with a therapeutic plateau, rather than a window, and suggest that in most cases there is no clinical advantage to the use of haloperidol doses greater than approximately 30 mg/day in schizophrenic patients.