The aim of this study was to determine the hormonal responses, which are dependent on the sleep wake cycle, to strenuous physical exercise. Exercise was performed after different nocturnal regimens: (i) a baseline night preceded by a habituation night; (ii) two nights of partial sleep deprivation caused by a delayed bedtime or by an early awakening; and (iii) two nights of sleep after administration of either a hypnotic compound (10 mg zolpidem) or a placebo. Eight well-trained male endurance athletes with a maximal oxygen uptake of 63.5 +/- 3.8 ml x kg(-1) x min(-1) (mean value +/- s(x)) were selected on the basis of their sleeping habits and their physical training. Polygraphic recordings of EEG showed that both nights with partial sleep loss led to a decrease (P< 0.01) in stage 2 and rapid eye movement sleep. A delayed bedtime also led to a decrease (P < 0.05) in stage 1 sleep. Zolpidem had no effect on the different stages of sleep. During the afternoon after an experimental night, exercise was performed on a cycle ergometer. After a 10-min warm-up, the participants performed 30 min steady-state cycling at 75% VO(2-max) followed by a progressively increased workload until exhaustion. The recovery period lasted 30 min. Plasma growth hormone, prolactin, cortisol, catecholamine and lactate concentrations were measured at rest, during exercise and after recovery. The concentration of plasma growth hormone and catecholamine were not affected by partial sleep deprivation, whereas that of plasma prolactin was higher (P < 0.05) during the trial after an early awakening. Plasma cortisol was lower (P < 0.05) during recovery after both sleep deprivation conditions. Blood lactate was higher (P < 0.05) during submaximal exercise performed after both a delayed bedtime and an early awakening. Zolpidem-induced sleep did not affect the hormonal and metabolic responses to subsequent exercise. Our results demonstrate only minor alterations in the hormonal responses to exercise after partial sleep deprivation.