Eight male cross-country runners and five male swimmers were tested four times during their collegiate seasons. Each trial corresponded to a different training load. The runners' trials were conducted before the start of organized practice (RT1), after 3 wk of increased training (RT2), 3 wk prior to the conference championship (pre-taper, RT3), and 4 d after the conference championship (post-taper, RT4). The swimmers' trials were conducted after the first 9 wk of training (ST1), after completing 2 wk of hard training (ST2), after an additional 6 wk of training (pre-taper, ST3) and during a week following the conference championship (post-taper, ST4). Venous blood samples, heart rate (HR) and blood pressure (BP) were obtained after 15 min supine rest (0700 h). Serum was analyzed for cortisol (C), total testosterone (TT), free testosterone (FT), and creatine kinase (CK). Blood samples (lactate), HR and RPE were obtained during a fixed velocity run (75% preseason VO2max) and blood samples and RPE following a 365.8 m swim (90% preseason VO2max). The runners then completed a "performance run" to exhaustion (110% preseason VO2max) and the swimmers completed maximal 22.9 and 365.8 m swims. Serum CK, C, TT, FT, and the TT:C and FT:C ratios were not significantly different among trials for the runners. Serum TT and FT were significantly (P < 0.05) lower for the swimmers at ST2 (TT 16.7 +/- 2.5; FT 85.3 +/- 8.5) compared to ST1 (TT 30.3 +/- 2.8; FT 130.2 +/- 20.9) whereas, C, TT:C or FT:C were not significantly altered.(ABSTRACT TRUNCATED AT 250 WORDS)