Anatomic factors associated with static posterior translation of the humeral head with or without glenohumeral osteoarthritis are unknown. We tested the hypothesis that there is an association between glenoid wear, glenoid version, and/or anteroposterior acromial tilt. Ninety-nine patients with glenohumeral joint degeneration involving advanced glenoid cartilage wear and/or rotator cuff disease scheduled for anatomic or reverse total shoulder replacement underwent standardized conventional radiographic and computed tomographic shoulder imaging. Measurements included glenoid version, humeral torsion, posterior acromial slope, and critical shoulder angle. The glenoid shape was classified according to Walch et al, and the integrity of the rotator cuff was assessed. Patients with glenoid type B2 or C had a median of 4° more glenoid retroversion (P = .022), a 5° less steep acromion (posterior acromial slope, 61° vs 56°; P = .004), and a higher combined score (glenoid version minus slope; odds ratio, 0.93 [95% confidence interval, 0.89-0.97]; P < .001; cutoff, -27°) than those with type A or B1. When the rotator cuff was torn, osteoarthritic changes were milder than when the cuff was intact (eg, P < .001 for supraspinatus). The study's hypothesis that the bony anatomy of the scapula and in particular the acromion is correlated with the type of glenoid wear was confirmed. Both a more horizontal acromial orientation in the sagittal plane and increased posterior glenoid version are found in osteoarthritis of the shoulder associated with eccentric, posterior glenoid wear. Tears of the rotator cuff are significantly associated with concentric osteoarthritis of the glenoid.