To test the hypotheses that short-term bed-rest (BR) deconditioning influences metabolic, cardiorespiratory, and neurohormonal responses to exercise and that these effects depend on the subjects' training status, 12 sedentary men and 10 endurance- and 10 strength-trained athletes were submitted to 3-day BR. Before and after BR they performed incremental exercise test until volitional exhaustion. Respiratory gas exchange and heart rate (HR) were recorded continuously, and stroke volume (SV) was measured at submaximal loads. Blood was taken for lactate concentration ([LA]), epinephrine concentration ([Epi]), norepinephrine concentration ([NE]), plasma renin activity (PRA), human growth hormone concentration ([hGH]), testosterone, and cortisol determination. Reduction of peak oxygen uptake (VO(2 peak)) after BR was greater in the endurance athletes than in the remaining groups (17 vs. 10%). Decrements in VO(2 peak) correlated positively with the initial values (r = 0.73, P < 0.001). Resting and exercise respiratory exchange ratios were increased in athletes. Cardiac output was unchanged by BR in all groups, but exercise HR was increased and SV diminished in the sedentary subjects. The submaximal [LA] and [LA] thresholds were decreased in the endurance athletes from 71 to 60% VO(2 peak) (P < 0.001); they also had an earlier increase in [NE], an attenuated increase in [hGH], and accentuated PRA and cortisol elevations during exercise. These effects were insignificant in the remaining subjects. In conclusion, reduction of exercise performance and modifications in neurohormonal response to exercise after BR depend on the previous level and mode of physical training, being the most pronounced in the endurance athletes.