Objective De-centering of the shoulder joint on radiographs is used as indicator for severity of rotator cuff tears and as predictor for clinical outcome after surgery. The objective of the study was to assess the effect of malrotation on glenohumeral centering on radiographs and to identify the most reliable parameter for its quantification. Subjects and methods. In this retrospective study (2014–2018), 249 shoulders were included: 92 with imaging-confirmed supra- and infraspinatus tears (rupture; 65.2 ± 9.9 years) and 157 without tears (control; 41.1 ± 13.0 years). On radiographs in neutral position and external rotation, we assessed three radiographic parameters to quantify glenohumeral centering: acromiohumeral distance (ACHD), craniocaudal distance of the humeral head and glenoid center (Deutsch), and scapulohumeral arch congruity (Moloney). Non-parametric statistics was performed. Results In both positions, only the distance parameters ACHD (< 0.5 mm) and Deutsch (< 1 mm) were comparable in the two study groups rupture and control. Comparing the parameters between the study groups revealed only ACHD to be significantly different with a reduction of more than 2 mm in the rupture group. Among the parameters, ACHD ≤ 6 mm was the only cut-off discriminating rupture (12–21% of the shoulders with ACHD ≤ 6 mm) and control (none of the shoulders with ACHD ≤ 6 mm). Ninety percent of shoulders with ACHD ≤ 6 mm presented with a massive rotator cuff tear (defined as ≥ 67% of the greater tuberosity exposed). Conclusion Glenohumeral centering assessed by ACHD and Deutsch is not affected by rotation in shoulders with and without rotator cuff tear. An ACHD ≤ 6 mm has a positive predictive value of 90% for a massive rotator cuff tear.