This study evaluated the reliability of the Constant-Murley Score. Two testers assessed 63 patients twice on the same day using the original publication by Constant and Murley. The intratester reliability of the total score was high and the differences between the tests were small; 2 of 14 items were unreliable. The intertester reliability was high, but there were significant median differences of the total score; 12 of 14 items were unreliable. We believe that the differences between the testers were due to the brief explanations of test components in the original publication. The reliability of the Constant-Murley Score could possibly be improved by a better standardization of the assessment procedure.