It has been suggested that, if dopamine antagonism is a necessary condition for the antischizophrenic action of neuroleptics, the prolactin response, as an index of dopamine blockade, would correlate with clinical response. Morning prolactin and clinical symptomatology were measured in 15 schizophrenic patients before neuroleptic therapy, and after three and six weeks of high-dose butaperazine or loxapine treatment. Prolactin levels were transiently elevated during the unmedicated admission period, probably reflecting a normal stress response. Prolactin increased in all patients during neuroleptic therapy. There was, however, no correlation between magnitude of prolactin changes and clinical response, probably because the prolactin response achieved a maximum at relatively low doses of neuroleptics.