Open-trial reports of substantial clinical improvement in most schizophrenic patients on hemodialysis for their psychiatric condition prompted the present study to determine the efficacy of hemodialysis under double-blind, controlled conditions. Fifteen schizophrenic outpatients were randomly assigned to either a real-sham or sham-real dialysis treatment sequence. Presented in detailed, graphic form, results of repeated measurement and other analyses of symptom and behavioral data collected initially, at crossover, and at the end of treatment revealed no differential effects between real and sham dialysis. These results provide important experimental evidence of the lack of therapeutic efficacy of hemodialysis in schizophrenia.