Male hyperprolactinemia (HPRL) is known to induce different types of sexual dysfunctions. In order to determine the incidence of HPRL among patients referred for sexual dysfunction, serum prolactin (PRL) was assayed in 1053 clinically idiopathic cases. Among 850 cases complaining of erectile impotence, 10 with marked HPRL (1.1%, PRL above 35 ng/ml) were found, of whom 6 cases were associated with a pituitary adenoma. 17 mild HPRL (2%, PRL 20–35 ng/ml) were also found. Among 124 cases with premature ejaculation, 13 (10%) mild HPRL were found. Serum PRL was normal in 51 cases complaining of anejaculation without orgasm, and 27 patients exclusively complaining of reduced sexual desire. Our results lay stress on the fact that serum PRL must be assayed in every case of clinically idiopathic erectile impotence. Indeed, 5 of the 10 marked HPRL patients would have been misdiagnosed if we had only assayed this hormone when plasma testosterone was below the normal range. Moreover, in order to shed some light on the mechanisms by which HPRL disturbs male sexual function, the sexual behaviour of 17 markedly HPRL males was compared to their serum levels of PRL and testosterone, first before treatment, then at regular intervals during treatment. Our main conclusion is that impotence cannot be totally explained by a decrease in plasma testosterone, because this steroid hormone was within the normal range in 7 of the 16 impotent patients. Moreover, when serum PRL was lowered by bromocriptine, 6 patients recovered their potency before plasma testosterone clearly increased, and in 3 of those patients before it reached the normal range. The other mechanisms able to explain the impotence of HPRL males are discussed. Our study suggests that psychological factors play an important additional role in the impotence of some HPRL patients.