Gait symmetry is often a goal for physiotherapy in patients with hip osteoarthritis and after total hip replacement. However, there is no agreement on criteria for pathological gait asymmetry. In this study we investigated discriminative abilities of trunk and footfall gait symmetry measures, and thereafter assessed whether a 10% cutoff value is valid as a general criterion of pathological gait asymmetry across measures. Anteroposterior, vertical, and mediolateral trunk symmetry, single support, and step length symmetry were obtained simultaneously by trunk accelerometry and an electronic walkway in 37 patients with end-stage hip osteoarthritis and 56 controls. Subjects walked six times along a 7-meter walkway at slow, preferred, and fast speed, before data were normalized for gait velocity. Anteroposterior, vertical, and single support symmetry measures showed best discriminating abilities. The general 10% criterion of gait asymmetry and optimal cutoff criteria calculated for each symmetry measure showed approximately equal total classification ability. However, the optimal cutoff criteria classified a high number of controls as having pathological gait asymmetry. The general criterion of 10% is valid with high total classification ability, does not classify asymmetry in able-bodied subjects as pathological, and is feasible for use on individual patients in the clinic as well as in research.