Intense exercise leads to accumulation of the inducible member of the 70-kDa family of heat shock proteins, Hsp70, in male, but not female, hearts. Estrogen is at least partially responsible for this difference. Because androgen receptors are expressed in the heart and castration leads to decreases in calcium regulatory proteins and altered cardiac function, testosterone (T) or its metabolites could also be involved. We hypothesized that removal of endogenous T production through castration would reduce cardiac Hsp70 accumulation after an acute exercise bout, whereas castrated animals supplemented with 5α-dihydrotestosterone (DHT) would show the intact male response. Fifty-four 8-wk-old male Sprague-Dawley rats were divided into intact, castrated, or castrated + DHT groups (n = 18/group). At 11 wk of age, 12 animals in each group undertook a 60-min bout of treadmill running at 30 m/min (2% incline) while the remaining 6 in each group remained sedentary. At 30 min or 24 h after exercise (n = 6/time point), blood and hearts were harvested for analysis. Serum T was undetectable in castrated and DHT-treated castrated rats, whereas serum DHT was significantly reduced in castrated animals only (∼60% reduction) (P < 0.05). Although there were no differences in constitutive levels of Hsp70 protein, exercise significantly increased cardiac hsp70 mRNA and protein in intact and DHT-supplemented rats, but not in castrated animals (P < 0.05). To examine whether castration eliminated the ability to respond to stress, another six intact and six castrated animals were subjected to a 15-min period of hyperthermia (core temperature raised to 42°C) and killed 24 h later. As opposed to exercise, castrated animals subjected to heat shock exhibited increases in Hsp70 above nonshocked (i.e., sedentary) animals, similarly to intact males (P < 0.05). These data suggest that androgens, in addition to estrogen, play a role in the sexual dimorphism observed in the stress response to exercise but not heat shock.