The aim of this study was to investigate the energy expenditure of walking with different aids in patients with spinal cord injury (SCI) and to compare the results with normal able-bodied control subjects. The tests were performed on a 20 m indoor path in a comprehensive rehabilitation clinic of a university hospital. Nine male patients with incomplete SCI at various levels from C6 to L2 and nine age- and gender-matched normal able-bodied subjects were enrolled in the study. Two different walking aids were evaluated: walker and crutch. The main outcome parameters--walking velocity, oxygen uptake and oxygen cost--were measured during ambulation with the two different walking aids. There was a statistically significant difference between controls and SCI patients in terms of walking velocity (P<0.001), and oxygen cost (crutches P<0.01; walker P<0.001). SCI patients walked more slowly and less efficiently. When we compared SCI patients' walking with the two different aids, a statistically significant difference was observed in velocity and oxygen cost (P<0.05) in favour of crutches. We concluded that energy expenditure studies are useful tools for giving objective measures to patients at the time of discharge and to encourage them to use efficient assistive devices in their daily activities.