We created unstable intertrochanteric fractures with a large posteromedial defect in eighteen pairs of fresh-frozen femora from cadavera, and used the femora to evaluate the stability of six combinations of treatment with sliding compression screws and sideplates. The variables that were tested were high (150-degree) angle plates compared with low (130-degree) angle plates, the position of the lag-screw in the femoral head and neck (in the center of the head as seen on both anteroposterior and lateral roentgenograms compared with posteroinferiorly), and whether or not medial cortical contact had been re-established with a limited osteotomy of the greater trochanter. The femora were loaded to the point of failure. The mode of failure, maximum load to failure, and bending rigidity of each method of fixation were measured. Of the six types of treatment, the use of a 150-degree-angle plate, position of the lag-screw in the center of the head as seen on both anteroposterior and lateral roentgenograms, and osteotomy of the greater trochanter resulted in the highest mean load to failure and the greatest rigidity. Over-all, re-establishment of medial cortical contact by means of an osteotomy of the greater trochanter significantly improved the mean load to failure and the rigidity of the fixation. The higher-angle plate, although more difficult to insert, increased the mean load to failure but had no effect on rigidity. The position of the lag-screw (in the center of the head as seen on both anteroposterior and lateral roentgenograms or posteroinferiorly) had no bearing on the mean load to failure, but it significantly affected the mode of failure.