The aim of this study was to compare the accuracy of four different methods for measuring wear using an apparatus that simulates known amounts of three dimensional wear. Wear was measured using the manual methods reported by Charnley, Livermore, Dorr and Wan and the computerized method reported by Devane. Only the method reported by Devane measured the three-dimensional (superior and anterior) wear with a reasonable accuracy, with a mean measurement error of 0.21 mm. With superior wear alone, Charnley's method underestimated the extent of wear by 16.6%, with a mean error of 0.35 mm; Livermore's method estimated wear to within 9.5%, with a mean error of 0.16 mm; Devane's method estimated wear to within 9.5%, with a mean error of 0.15 mm; and Dorr's method underestimated wear by 25.4%, with a mean error of 0.56 mm. Dorr's method was modified as a result of the experimental tests. The clinical application of the new method showed comparable data to that using the Devane method. In conclusion, this new method can be used to estimate the average wear in groups of patients accurately.