Prolactin elevation is both a common and a persistent event with the currently marketed antipsychotics, excluding clozapine. Elevations have been associated with both acute (galactorrhea, amenorrhea) and chronic (predisposition to osteoporosis) treatment-emergent adverse events. One of the defining criteria for an atypical antipsychotic is the relative lack of persistent prolactinemia. A double-blind, placebo- (N = 68) and haloperidol- (Hal: 15 +/- 5 mg/day, N = 69) controlled trial of three dose ranges of olanzapine (Olz-L: 5 +/- 2.5 mg/day, N = 65; Olz-M: 10 +/- 2.5 mg/day, N = 64; Olz-H: 15 +/- 2.5 mg/day, N = 69) in the treatment of schizophrenia afforded the opportunity to assess the temporal course of the influence of olanzapine and haloperidol on serum prolactin concentration. Consistent with its potent D2 antagonism, haloperidol was associated with a statistically significantly higher incidence of treatment-emergent prolactin elevation (72%) than seen with placebo (8%; p < 0.001) at week 2 of therapy. Expectedly, this elevation was also persistent at weeks 4 and 6. In contrast, olanzapine-associated treatment-emergent prolactin elevations were both lower in magnitude and transient. At week 2, 38% of the Olz-H, 24% of the Olz-M, and 13% of the Olz-L treatment groups exhibited a treatment-emergent prolactin elevation, with a mean increase of 0.35, 0.52, and 0.61 nmol/l, respectively; for haloperidol the mean increase was 1.23 nmol/l. For only the Olz-M and the Olz-H treatment groups did the week 2 incidence of treatment-emergent prolactin elevations differ statistically significantly from placebo. Both the incidence of elevations and the mean increase, in prolactin concentration were less than that seen with haloperidol. Furthermore, by treatment week 6, all three olanzapine groups exhibited incidences of treatment-emergent prolactin elevation that were comparable to placebo and were statistically significantly less than observed with haloperidol. Rapid adaptation was observed in the temporal course of prolactin elevations associated with olanzapine based on both the categorical analysis of treatment-emergent high values and the analyses of temporal change in mean concentrations. In contrast to haloperidol, the magnitudes of the treatment-emergent elevations associated with olanzapine were minimal. The rates of elevation were approximately one-half to one-third those observed with haloperidol and were significantly more transient. Olanzapine, even at the highest doses (15 +/- 2.5 mg/day) used, was not associated with persistent elevations of prolactin, consistent with an 'atypical' pharmacologic profile.