Selective face ventilation, known to interfere with bloodstreamed temperature signals, should alter exercise-induced prolactinemic (PRL) responses if, as recently suggested, the latter responses are associated with body temperature increases generated by working muscles. To test this hypothesis, 15 trained adult male and female volunteers were submitted to 45-min hyperthermic treadmill runs with and without selective face ventilation. Tympanic (Tt) and rectal (Tr) temperatures were measured in these experiments, and serum immunoreactive PRL was assayed on venous blood sampled immediately before and immediately after the exercise. Hyperthermic running induced significant Tr elevations whether selective face ventilation was present or not. In the absence of face ventilation, serum PRL rose markedly in 11 of the 15 hyperthermic runners, the lack of response in 4 runners confirming that exercise per se is not necessarily hyperprolactinotrophic. When the 11 responders (11/15) ran with face ventilation, serum PRL response was abolished in 6 of them, indicating a significant interference from face ventilation on temperature signals streamed through carotid blood; serum PRL-maintained responses in 5 of these 11 ventilated runners suggest the absence of a learning effect. It is concluded that the blood-streamed temperature signals arising from working muscles represent an important but not a unique determinant of blood PRL response during aerobic running. The contribution of TRH and plasma volume changes to blood PRL variations during hyperthermic running appeared nonsignificant.