Supraspinatus outlet view and thirty-degree caudal tilt view have been used for the diagnosis of rotator cuff disease. Because scapula is moving on the thoracic cage, these two views are not always easy to take. The purposes of this study were to investigate the detectability of these views, to compare both views and to observe the interobserver reliability of these views. Supraspinatus outlet views and thirty-degree caudal tilt views were taken from 66 shoulders with clinically diagnosed rotator cuff disease. The shape of the acromions was examined on the supraspinatus outlet views and the size and shape of the acromions were analyzed on the thirty-degree caudal tilt views. An orthopaedic resident, an orthopaedic staff and a radiologic staff reviewed the all films without knowing the opinions of others. Suprapinatus outlet views were not-acceptable in average 43.9% and thirty-degree caudal tilt view were not-acceptable in average 12.6%. All of the three observes had same opinions in 43.9% for the supraspinatus outlet views, and in 45.5% for in thirty-degree caudal tilt views. Disagreement among all three observes was noticed in 13.6% for the supraspinatus outlet views and in 3.0% for in thirty-degree caudal tilt views. In the supraspinatus outlet views, the hooked acromions appeared in average 12.1%. In the thirty-degree caudal tilt view, acromial spurs appeared in average 27.3%. Only 2.5% in average showed hooked acromion on supraspinatus outlet view and acromial spur on thirty-degree caudal tilt view. Thirty-degree caudal tilt view is easier to take and more reliable than supraspinatus outlet view. In is better to take both supraspinatus outlet view and thirty-degree caudal tilt view for each patients with rotator cuff diseases. Radiologic findings should be judged on the base of clinical feature.