Hormone and substrate responses to mild and heavy treadmill exercise were compared in women who used oral contraceptives (OC group; n = 7) and in normally menstruating women (control group; n = 8). Venous blood samples were obtained before exercise (-5 min), during exercise (15, 30, 45, and 60 min), and 30 min after exercise. All samples were analyzed for glucose, lactate, free fatty acids (FFA), glycerol, follicle-stimulating hormone (FSH), luteinizing hormone (LH), human growth hormone (hGH), cortisol, insulin, estradiol (E2), and progesterone (P). Substrate patterns during exercise were not altered by the phase of the menstrual cycle or OC usage. However, in the OC group the FFA concentrations were consistently higher during mild exercise and the glucose concentrations were lower at rest and during exercise than in the control group (P less than 0.05). No differences in lactate or glycerol responses were observed between the groups (P greater than 0.05). The responses of insulin and hGH to exercise were not related to the OC use per se but rather to the steroid status, either endogenous or exogenous. Specifically, during the steroid phases (OC use phase and luteal phase) 1) insulin concentrations were not quite as markedly reduced (i.e., 12% higher when luteal phase and OC usage phase data were combined; P less than 0.05), and 2) hGH concentrations at rest and during light exercise were higher in the OC group during the OC use phase (P less than 0.05). LH patterns were not affected by exercise (P greater than 0.05), but a slight decrease was found in FSH (P less than 0.05). Increments in P and E2 were observed in the control group in both the follicular and luteal phase (P less than 0.05), but much greater increments in P occurred in the luteal phase than in the follicular phase (P less than 0.05). In contrast to the control group, no increments in P, E2, or cortisol occurred in the OC users during exercise (P greater than 0.05). Therefore the new observations in this study are that 1) insulin and growth hormone respond in a complex manner during exercise with either the phase of the menstrual cycle or the phases of OC use and disuse and 2) the steroid concentrations (P, E2, cortisol) are increased in the controls but not in the OC users during exercise. The latter point suggests that normal steroid increments are due to an increased rate of secretion rather than a decrease in the hepatic clearance of these steroids.(ABSTRACT TRUNCATED AT 400 WORDS)