Increases in step width have been reported for several clinical populations, including older adults and stroke survivors. These populations often also exhibit decreased hip abductor strength, suggesting that walking with wider steps may be an adaptive response in order to reduce the mechanical demands on the hip abductors. The purpose of this study was to quantify the relationship between step width and gluteus medius (GM) activity during walking. Fourteen young, uninjured adults walked on a treadmill at 1.25 m/s for four step width conditions (Normal, Narrow, Medium, and Wide) while step width and stance phase GM electromyographic (EMG) activity were quantified. We also measured hip abduction torque and GM activity during maximum voluntary isometric contractions (MVICs) at three hip angles (neutral, abducted 10°, and abducted 20°). During walking trials, GM activity was significantly (p < 0.0001) influenced by step width; compared to Normal walking, GM activity was 47% higher with Wide steps and 24% lower with Narrow steps. We also observed a weak positive correlation (r = 0.18 ± 0.14) between step width and GM activity during Normal walking, as GM activity was higher with wider steps. These results cannot be attributed to changes in GM conformation under the recording electrode, as GM activity was not influenced by hip angle during MVICs. The increased GM activity with wider steps does not support the proposal that increasing step width would be a beneficial adaptation to weakened hip abductors. A likely alternative explanation is that increased step width is a response to decreased gait balance.