The present study aimed to investigate intraindividual changes in plasma prolactin concentrations by switching from haloperidol treatment to risperidone treatment. The subjects were 15 female schizophrenic inpatients who received firstly haloperidol 12 mg/day for at least 2 weeks and, thereafter, risperidone 6 mg/day. Prolactin concentration in plasma during risperidone treatment (median 87.5 ng/ml, range 5.3-298.1 ng/ml) was significantly ( P<0.01) higher than during haloperidol treatment (median 50.7 ng/ml, range 11.6-226.6 ng/ml). In contrast, the ratio of prolactin concentration to nmol/l unit drug concentration (the active moiety: the sum of risperidone and 9-hydroxyrisperidone) during risperidone treatment (median 1.10, range 0.02-3.73) was significantly lower ( P<0.001) than that of haloperidol (median 1.81, range 0.41-8.24). Prolactin concentrations during both treatment phases correlated well in individuals (r=0.619, P<0.05), whereas better correlation was found in the ratio of prolactin concentration to drug concentration (r=0.779, P<0.01). These findings suggest the higher risk of hyperprolactinemia during risperidone treatment than during haloperidol treatment at clinically used dosages. However, from a purely pharmacological point of view, prolactin response per drug concentration was more sensitive during haloperidol treatment than risperidone treatment, probably resulting from the potent and selective antagonistic effect of haloperidol on dopamine D(2) receptor, compared with the broader pharmacological spectrum of risperidone.