We examined the relationship between physical fitness and circulating components of the GH-insulin-like growth factor I (IGF-I) system [i.e. GH, GH-binding protein (GHBP), IGF-I, and IGF-binding proteins 1-5 (IGFBP-1 through-5)] in adolescent females (age range, 15-17 yr). The study consisted of 1) a cross-sectional protocol (n = 23) in which GH-IGF-I components were correlated with fitness, as estimated by thigh muscle volume and maximal O2 uptake; and 2) a prospective study in which fitness, GH-IGF-I system components, and osteocalcin were examined before and after a 5-week period of endurance-type training (control, n = 6; trained, n = 10). The cross-sectional analysis revealed significant (P < 0.05) positive correlations between fitness and 1) mean 12-h overnight GH levels, 2) GHBP, and 3) IGF-I. Muscle volume was negatively correlated with both IGFBP-2 and -4. The prospective training study was associated with 1) increases in circulating osteocalcin (39 +/- 14%; P < 0.007), and 2) decreases in IGF-I (-14 +/- 5%; P < 0.05) and IGFBP-5 (-10 +/- 4%; P < 0.04). Unexpectedly, IGFBP-3 fell in both control (-8 +/- 2%; P < 0.01) and trained subjects (-5 +/- 3%; P < 0.05), and GHBP was reduced only among control subjects (-10 +/- 7%; P < 0.04). In summary, fitter adolescent girls tended to have increased mean serum GH, GHBP, and IGF-I. In contrast, brief endurance training led to increases in muscle mass and serum osteocalcin that were not accompanied by increases in GH or IGF-I. In fact, training may, in the short term, have led to a catabolic state hormonally expressed by reductions in IGF-I and IGFBP-5.