To test the hypothesis that bed rest leads to larger decrease in maximal oxygen consumption (VO_2max) upright (U) than supine (S), because of adequate cardiovascular response in S, but not in U, VO_2max and maximal systemic oxygen delivery (QaO₂) were determined during graded exercise on a cycle ergometer on 10 healthy subjects. At each power, VO₂ was determined breath-by-breath and VO_2max was established from the plateau in the relationship between steady state VO₂ and power. Blood pressure and heart rate (fh) were recorded beat-by-beat. Stroke volume (Qst) was determined by Modelflow from pulse pressure profiles. Cardiac output (Q) was obtained as fh times Qst. Arterial oxygen concentration (CaO₂) was computed from haemoglobin and arterial oxygen saturation measurements. QaO₂ was obtained as Q times CaO₂. Before bed rest,VO_2max was the same in U and S. After bed rest, it was reduced in both U and S (-38.6% and -17.0%, respectively), resulting 30.8% higher in S than in U. Maximal fh was the same in all conditions. Maximal Qst was reduced after bed rest in U (- 44.3 %), but unchanged in S (+ 3.7 %), resulting 98.9 % higher in S than in U. Maximal Q was equal in U and S before bed rest. After bed rest, it decreased in U (- 45.1%), but not in S (+ 9.0 %), resulting higher in S than in U (+ 98.4 %). QaO₂ was reduced after bed rest in U (- 37.8%), but not in S (+ 14.8 %), being higher (+ 74.8%) in S than in U. We conclude that i) the VO_2max decrease after bed rest in S is not due to cardiovascular alterations, and thus it depends on peripheral oxygen transfer limitation only; ii) the lower VO_2max after bed rest in U than in S is due to cardiovascular response in U; iii) the VO_2max decrease after bed rest in U is the result of the interaction of impaired cardiovascular response and limited peripheral oxygen transfer. The cardiovascular system is preserved in microgravity, but upon gravity resumption, it provides inadequate response to exercise in U.