The objective was to define the diagnostic validity of prolactin response tests by comparing the stimulating effect of thyrotropin-releasing hormone (TRH) and metoclopramide on prolactin secretion in patients with and without functional hyperprolactinemia. Prolactin response studies were performed in 18 patients with functional hyperprolactinemia (defined as prolactin serum levels > or = 16 and < 50 ng/ml during the follicular phase without evidence of prolactinoma) and 18 controls with similar age who had normal serum prolactin levels. Tests were done on the 7th, 8th or 9th day of the follicular phase under standardized conditions (at 08.00-09.00, after a 1-h rest and overnight fasting) with 200 micrograms TRH or 10 mg metoclopramide i.v. After metoclopramide, higher maximal prolactin levels were observed in the study group as well as in the control group than after TRH (metoclopramide mean: 243 +/- 62 ng/ml vs. 181 +/- 100 ng/ml, U-test: p = 0.0019; TRH mean: 101 +/- 23 ng/ml vs. 41 +/- 20 ng/ml, U-test: p = 0.0001). There was no significant difference in the relative increment of prolactin serum levels between both groups, neither after metoclopramide nor after TRH (after metoclopramide, mean: 13.5 +/- 5.8 vs. 17.2 +/- 9.2, U-test: p = 0.137; after TRH, mean 3.6 +/- 1.1 vs. 3.8 +/- 1.9, U-test: p = 0.899). No correlation between prolactin basal and peak levels, either after TRH or after metoclopramide, could be found, neither in the study (metoclopramide test: p = 0.738; TRH test: p = 0.076) nor in the control group (metoclopramide test: p = 0.331; TRH test: p = 0.360). While significantly higher prolactin peaks were recorded after metoclopramide than after TRH in both groups, no difference in the response to TRH and metoclopramide, regarding absolute and relative increment, could be found.