BACKGROUND: Aerobic exercise capacity provides information regarding cardiorespiratory health and physical capacity. However, in many populations the ability to measure whole-body or leg aerobic exercise capacity is limited due to physical disability or lack of appropriate equipment. Clinically there is a need to evaluate aerobic capacity in individuals who cannot use their legs for locomotion. In astronauts the habitable space for exercise testing in the next generation of space exploration systems may be restricted and may not support the traditional lower body testing. Therefore, the purpose was to determine if upper body physical performance could estimate lower body aerobic capacity. METHODS: Maximal O-2 uptake ((V) over dot(O2max)), gas exchange threshold (GET), and the highest sustainable rate of aerobic metabolism [arm cranking critical power (ACP) and lower body critical speed (LCS)] were determined in 55 conditioned men and women during arm-cranking and treadmill running. RESULTS: (V) over dot(O2max) and GET (48.6 +/- 7.6 and 29.0 +/- 4.8 ml . kg(-1) . min(-1), respectively) were significantly lower during arm-cranking exercise compared to running (27.1 +/- 7.6 and 13.5 +/- 2.6 ml . kg(-1) . min(-1), respectively). The (V) over dot(O2) at ACP was significantly lower than the (V) over dot(O2) at the LCS (18.4 +/- 5.01 vs. 39.5 +/- 8.1 ml . kg(-1) . min(-1), respectively). There was a significant correlation between arm-cranking and lower body (V) over dot(O2max), GET, and the (V) over dot(O2) at LCS and ACP. Backward stepwise regression analyses revealed that arm-cranking physical fitness could explain 67%, 40%, and 49% of the variance in lower body (V) over dot(O2max), GET, and LCS, respectively. DISCUSSION: Results suggest arm-cranking exercise can be used to obtain an approximation of lower body aerobic capacity.