The potential importance of neuroleptic activity measures in the management of schizophrenia warrants attention to the methods for assessing neuroleptic bioactivity and stability of neuroleptic bioactivity over time. We have carried out measurements of serum neuroleptic and prolactin concentrations in 18 schizophrenic patients treated with haloperidol or thioridazine for up to 1 year. Serum neuroleptic levels were measured by a radioreceptor assay using porcine striatum. The lower limit of sensitivity of the assay was 0.6 ng haloperidol/ml, the intra- and interassay coefficients of variation 3 and 9%, respectively. A linear correlation was observed between haloperidol dose (5-30 mg/d) and serum neuroleptic activity (r = 0.706, P less than 0.001) and a curvilinear relationship between thioridazine dose (50-600 mg/d) and serum neuroleptic activity in schizophrenic outpatients. There was a positive correlation between serum neuroleptic and prolactin concentrations for the patients taking haloperidol (r = 0.620, P less than 0.001) or thioridazine (r = 0.542, P less than 0.001). In patients taking a constant dose of haloperidol or thioridazine for up to 1 year serum neuroleptic activity remained stable, suggesting the absence of metabolic tolerance; the observation of a decrease of 38 +/- 16% (mean +/- SD) in serum prolactin concentrations in patients treated with haloperidol but no prolactin decrease with thioridazine suggests that under certain neuroleptic treatment conditions a functional tolerance develops in the tuberoinfundibular dopamine system.