METHODS Randomized controlled trail. OBJECTIVE To investigate if people with incomplete spinal cord injury (SCI) can perform high-intensity weight-bearing exercise by comparing cardiovascular responses at maximal workloads during stationary cycling and treadmill walking, and to explore mechanical efficiencies at sub-maximal workloads. METHODS Sunnaas Rehabilitation Hospital, Nesoddtangen, Norway. METHODS Fifteen people with incomplete SCI and 15 healthy control subjects performed sub-maximal and maximal exercise tests of both stationary cycling and uphill treadmill walking on separate days. Oxygen uptake (VO(2); l min(-1) and ml kg(-1) min(-1)), carbon dioxide production (VCO(2); l min(-1)), respiratory exchange ratio (RER) and heart rate (HR) were continuously measured throughout the tests. RESULTS The SCI group showed no significant differences in peak VO(2) (2.42±0.68 l min(-1) versus 2.58±0.76 l min(-1), P=0.19) or other cardiovascular responses at maximal workloads for stationary cycling as compared with uphill treadmill walking, except for higher RER during the cycle test. The control subjects exhibited a significantly higher peak VO(2) during the treadmill test as compared with the cycle test (P=0.007). Both groups had lower mechanical efficiency when walking as compared with cycling, but the mean difference between cycling and walking was not significantly different between the groups during sub-maximal workloads (P >0.24). CONCLUSIONS Subjects with incomplete SCI were able to perform high-intensity weight-bearing exercise and exhibited similar mechanical efficiencies at sub-maximal workloads as healthy controls. Uphill walking might be a good alternative to weight-bearing exercise for increasing the physical capacity of people with incomplete SCI.