The purpose of this study was to examine serum and salivary cortisol responses to cycling exercise in male children, 10.6 +/- 0.2 yr (mean +/- SE). Each child performed a graded exercise test on a cycle ergometer to determine VO2max. On a separate day, a 30-min bout of exercise at 70% of VO2max was performed. Blood, obtained from a venous catheter, and saliva samples were collected at rest, at 15 and 30 min of exercise, and 15 min post-exercise. The mean serum cortisol level at 15 min (7.94 +/- 1.43 micrograms.dl-1) and 30 min (8.72 +/- 1.77 micrograms.dl-1) of exercise and at 15 min post-exercise (8.21 +/- 1.59 micrograms.dl-1) were significantly greater than rest (5.54 +/- 0.86 micrograms.dl-1). The increase in salivary cortisol levels over time approached (P = 0.08), but did not reach significance. However, effect size analyses indicated that the increase in salivary cortisol at 30 min of exercise (0.64) and 15 min post-exercise (0.62) was similar to the change in serum cortisol at these same two time points (0.72 and 0.66, respectively). Serum and salivary cortisol were correlated (P < 0.05) at 15 min of exercise (r = 0.77), 30 min of exercise (r = 0.90), and 15 min post-exercise (r = 0.84), but not at rest (r = 0.46). In conclusion, 30 min of submaximal exercise at 70% of VO2max significantly increased serum cortisol level; and salivary and serum cortisol are correlated during and after exercise.