Visceral movement due to impact loading is believed to play a role in the locomotor-respiratory coupling (LRC) that has been detected in a number of mammalian species. In the bird and bat species in which LRC has been described, the effect of the wing muscles on the timing of respiration appears to be a dominant influence. To test the hypothesis that LRC occurs in humans propelling wheelchairs (where there is no impact loading and the arms are used for locomotion), we studied 10 wheelchair athletes on a motorized treadmill at three speeds. Each subject's data were analyzed by spectral analysis (based on the fast Fourier transform), which detected apparent LRC (rates within 1% of a single-digit integer ratio) in 12 (40%) of the 30 test settings. However, a control analysis, in which each subject's arm-thrust rates were compared with another subject's breathing rates, revealed apparent (but false) coupling in 8 (27%), not significantly less often (using the chi 2 test). These findings appear to refute the hypothesis that LRC occurs during wheelchair propulsion. These data are consistent with the theory that the visceral piston is important to LRC and suggest that rhythmic arm movements are insufficient to induce the phenomenon in this setting.