BACKGROUND A large critical shoulder angle (CSA) >35° is associated with the development of rotator cuff tearing. Lateral acromioplasty (AP) has the theoretical potential to prevent rotator cuff tearing and/ or to reduce the risk of re-tears after repair. It is, however unclear which part of the lateral acromion has to be reduced to obtain the desired CSA. It was the purpose of this study to determine which part of the lateral acromion has to be resected to achieve a desired reduction of the CSA in a given individual. METHODS First, the influence of the exact radiographic projection on the CSA was examined. Second, the influence of anterolateral versus strict lateral AP on the CSA was studied in eight scapulae with different anatomic characteristics. Differences in CSA reduction were investigated using paired t-test or Wilcoxon test. RESULTS Scapular rotation in the sagittal and axial plane had a marked influence on the radiologically measured CSA ranging from -6 to +16°. Overall, lateral AP of 5/10mm reduced the CSA significantly greater than anterolateral AP of 5mm/10mm [5mm: 2.3° (range: 0.7°-3.6°) SD±0.8° vs. 1.2° (range: 0°-3.3°) SD±1.1°, p=0.0002]/[10mm: 4.8° (range: 2.1°-7°) SD±1.3° vs. 2.7° (range: 0°-5.3°) SD±1.7°, p=0.0001]. Depending on scapular anatomy anterolateral AP did not alter CSA at all. CONCLUSIONS For comparison of pre- and postoperative CSA, the exact orientation of the X-ray and the spatial orientation of the scapula must be as identical as possible. Anterolateral AP may not sufficiently correct CSA in scapulae with great acromial slopes and smaller relative external rotation of the acromion as the critical acromial point (CAP) may be located too posteriorly and thus is not addressed by anterolateral acromioplasty. Consistent reduction of the CSA could be achieved by lateral AP in all eight scapulae.