This study attempts to apply geometric morphometric techniques for the analysis of 3D kinematic marker-based gait data. As a test, we attempted to identify sexual dimorphism during the stance phase of the gait cycle. Two techniques were used to try to identify differences in the way males and females walk without the results being affected by individual differences in body shape and size. Twenty-eight kinematic markers were placed on the torso and legs of 6 male and 8 female subjects, and the 3D time varying coordinates of the kinematic markers were recorded. The gait cycle trials were time-normalized to 61 frames representing the stance phase of gait, and the change in the shape of the configuration of kinematic markers was analyzed using principal components analysis to produce ‘gait signatures’ that characterize the kinematics of each individual. The variation in the gait signatures was analyzed with a further principal components analysis. These methods were able to detect significant sexual dimorphism even after the effects of sexual body shape and size differences were factored out. We discuss insights gained from performing this study which may be of value to others attempting to apply geometric morphometric methods to motion analysis.