Human gait analysis has become an active field of research with applications in clinical diagnosis, monitoring, and rehabilitation. A three-dimensional motion capture system enables researchers to obtain high accuracy in gait identification. Competitive swimming is one of the most popular sports among children in the U.S, and competitive swim training requires repetitive motions of the lower body that may result in musculoskeletal changes that generate different gait patterns. However, whether repetitive motions during swim training affect gait in children is unknown. Here, gait patterns of nine competitive swimmers aged 5–12 years were compared with age-matched nine controls who do not swim. A motion capture and analysis system and quantitative analysis methods were used to examine spatiotemporal, kinematic, and kinetic parameters. The mean maximum pelvic tilt of swimmers was approximately half that of controls (P = 0.001). The maximum hip flexion in both stance (P = 0.009) and swing (P = 0.040) gait phases was also less, whereas the mean maximum knee flexion (P = 0.022) was greater in swimmers. The average maximum hip angle was less in swimmers (P = 0.041). Hip and knee joint moments showed no difference, but concentric ankle power (P = 0.013) and plantarflexion moment (P = 0.050) were less in swimmers. Thus, while the range of motion of the knee was greater in swimmers, those of the hip and pelvis were substantially less.