The purpose of this study was to compare the humeral retroversion measured using a CT with that using a radiogram. We analyzed twenty healthy volunteers (forty humeri). Ten were male and ten were female. We adopted two measuring methods using the CT to draw the humeral head axis; Simeonides' Method (Method S)and our original method (Method A). The cubital axis was drawn as the tangent to the epicondyles. We also adopted a modified Soderlund's Method in the measurement using a radiogram. In his original method, it is impossible to measure the humeral retroversion because we can not define the edge of the articular surface of the humerus. The retroversion angle using Method S was 34.8 ± 12.9°, and using Method A was 32.6 ± 10.7°. The result of the measurement based radiogram was 28.9 ± 10.6°. The mean value of the retroversion measured using CT was larger than that of using a radiogram. There was a significant difference of the mean value of the retroversion between the result of the measurement using CT and radiogram. The difference of the humeral retroversion between Method S and A was not significant. We also analyzed the humeral retroversion in 10,15,20, and 30 of abduction of the shoulder joint with the radiogram based method. The retroversion angle tended to decrease gradually with the increase of the abduction angle. The results of this study were different from Soderlund's report.
