The authors gave 48 newly admitted schizophrenic patients an initial test dose (2.2 mg/kg) of chlorpromazine hydrochloride and additional doses at 24 and 36 hours. Clinical response was measured at 48 hours and at the end of subsequent treatment with a fixed dose (6.6 mg/kg) of chlorpromazine. There was a consistent correlation between 48-hour change and the eventual amount of improvement by the end of drug treatment for all 10 outcome criteria. The authors conclude that change early on in drug treatment is a useful predictor of outcome.