To clarify the influence of body position on exercise prescription, 14 men (mean age +/- standard deviation 60.0 +/- 6.1 years) with coronary artery disease who underwent randomized recumbent and upright cycle ergometer tests to volitional fatigue were studied. At 100 watts, heart rate (HR), systolic blood pressure, oxygen consumption (VO2), rate pressure product and rating of perceived exertion were greater (p less than 0.05) in the upright than in the recumbent position. At peak exercise, however, these variables were not significantly different. Regressions of relative HR versus VO2 for recumbent and upright cycle ergometry were comparable: y = 1.24x - 32.7 and y = 1.26x - 31.5, respectively, where y = % maximal VO2, and x = % maximal HR. These findings indicate that recumbent exercise prescriptions may be based on the peak HR and VO2 values obtained during upright cycle ergometry, and vice versa. However, differences in the cardiorespiratory responses at submaximal exercise preclude the interchangeability of upright and recumbent training work rates.