This study examined the possibility that fatigue may modify the hormone responses to exercise. A group of 12 endurance trained athletes ran for 2 h (blood lactate concentrations of approximately 2 mmol x l(-1)) in order to induce fatigue. The subjects exercised for 10 min at 70% maximal oxygen uptake before (1st test) and after (2nd test) the 2 h run to assess hormone responsiveness. A 1 min anaerobic power test was performed to assess muscle power. Cortisol, growth hormone, testosterone and insulin concentrations were determined before and after the 1st and 2nd tests. The 1st test resulted in increases in concentrations (P < 0.05) of cortisol and growth hormone, a decrease in insulin concentration (P<0.01) and no change in testosterone concentration. The 2 h run caused decreases of insulin, increases of growth hormone concentration and variable responses in the concentrations of cortisol and testosterone. The 2nd test decreased insulin concentration further (P < 0.05), but responses of the concentrations of testosterone, growth hormone and cortisol were variable. In 6 subjects (group A) cortisol displayed an increase [mean (SD)] from baseline concentrations [+ 304.0 (60.0) nmol x l(-1)], while in the other 6 subjects (group B) a decrease or no change was seen [+ 3.1 (5.3) nmol x l(-1), between groups, P<0.05]. Growth hormone concentration was substantially higher in group A [+ 14.7 (4.8) ng x ml(-1)] than group B [+ 6.0 (2.9) ng x ml(-1)] following the 2nd test. In group A anaerobic muscle power was higher, while in group B it was lower, after the 2 h run than before the 2 h run (P < 0.05). The findings suggest that fatigue from prolonged endurance activity may introduce a resetting in the pituitary-adrenocortical component of the endocrine system, expressed either by intensified or by suppressed endocrine functions.