Preliminary findings are reported for a study of the relative efficacy of conventional treatment alternatives routinely used to treat acutely relapsed schizophrenic patients who fail an initial course of standard neuroleptic therapy. A sample of 156 acutely ill schizophrenic, schizoaffective, and schizophreniform patients who had been hospitalized recently in an acute-care inpatient facility were treated openly with fluphenazine (FPZ) 20 mg/day and with prophylactic benztropine for 4 weeks. Those subjects who failed to meet a priori criteria for substantial therapeutic response at the end of Week 4 were randomized to receive double-blind treatment for an additional 4 weeks with one of the following: (1) the same dose of neuroleptic (FPZ 20 mg/day); (2) an increased dose of neuroleptic (FPZ 80 mg/day); or (3) a different class of neuroleptic (haloperidol 20 mg/day). Of the 115 subjects who completed the open phase of study, 32 percent were identified as responders. Higher negative symptom scores and increased acute extrapyramidal side effect ratings appeared to distinguish the nonresponder from the responder group. Of the nonresponders who went on to randomized treatment, only 4 of 47 subjects (9%) subsequently responded. No superior efficacy was associated with any of the specific alternative treatments studied.