Six hundred forty newly-admitted schizophrenic men in 35 VA hospitals were randomly assigned to chlorpromazine, triflupromazine, mepazine, prochlorperazine, perphenazine and phenobarbital groups. Treatment followed a double blind procedure for 12 weeks. Patients were started on low "equivalent" doses of each drug which were gradually increased in a predetermined manner during the first 4 weeks. During the final 8 weeks, each prescribing physician adjusted the dose for each of his patients in order to evoke an optimal therapeutic response.Average daily doses during the flexible period were: chlorpromazine, 635 mg.; triflupromazine, 175 mg.; mepazine, 190 mg. ; prochlorperazine, 90 mg. ; and perphenazine, 50 mg. Clinical evaluations using two rating scales provided 24 criteria of change. For each criterion, the mean of each of the 6 treatment groups adjusted for the net effect of 12 control variables was compared by analysis of multiple covariance with the mean of every other treatment group at each of three evaluation periods; first month, the following 2 months, and over the entire 3 months. Side effects, hematologic and hepatic function data were also recorded during the course of treatment. One hundred sixty-eight patients failed to complete the study.In general, the results indicated that all 5 phenothiazine derivatives were therapeutically more effective than phenobarbital. Mepazine was less effective than the other 4 drugs at the doses employed. No significant differences in therapeutic efficacy were noted between chlorpromazine, triflupromazine, prochlorperazine, and perphenazine. Criterion measures showing change toward improvement after treatment with phenothiazine derivatives included resistiveness, belligerence, thinking disturbance, and degree of illness. Other criteria affected favorably, especially by the 4 more potent phenothiazines, were motor disturbance, paranoid projection, perceptual distortion and withdrawal.Only 21 patients (3%) were discontinued from treatment because of side reactions or deviant laboratory tests. Most side reactions, especially the extrapyramidal syndromes, were produced by perphenazine and prochlorperazine. Phenobarbital was associated with a number of side reactions ("turbulence," autonomic symptoms) commonly attributed only to the phenothiazine derivatives. Abnormal hematologic tests including eosinophilia, leucocytosis and leucopenia were neither frequent nor severe. The distribution of the 36 patients with leucopenia was not significantly different among the treatment groups. Continued treatment with the drugs in 31 leucopenic patients produced no case of agranulocytosis. Although abnormal hepatic tests occurred in 88 patients, these were sporadic. No clearcut case of jaundice or hepatic dysfunction was encountered during treatment.Abstract Teaser