The development, validation, and application of a new radioimmunoassay for haloperidol in biological fluids is described. The antiserum, raised against N-amino-butyl chlorophenyl piperidine bovine albumin conjugate, could not distinguish between haloperidol and its reduced metabolite, but it could discriminate against chlorophenyl piperidine (cross-reaction 2.6%). The fluorophenyl metabolites of haloperidol were not recognized by the antiserum. Haloperidol determinations were made on less than 100 microliter aliquots of human and rhesus monkey plasma or diluted urine without prior extraction of the sample. The radioimmunoassay was applied to the study of the pharmacokinetics of intravenous haloperidol administration to two male rhesus monkeys. Salient features of the results are as follows. As with man, the plasma concentration versus time curve could be resolved into three compartments, but there were differences in the distribution of haloperidol between the compartments. The apparent volume of distribution for the two monkeys examined was 5.87 L kg-1 and 7.37 L kg-1, considerably smaller than in man, a difference almost entirely due to a much smaller tissue compartment. The biological half-life of 15.97 hr and 7.56 hr was similar to man. The mean hepatic extraction ratio was calculated to be 0.032 and 0.056, and the data suggested that hepatic metabolism of haloperidol may be of lesser importance in rhesus monkey than in man. An insignificant proportion (0.01%) of the administered dose was excreted as haloperidol in the urine.