BACKGROUND: The utility of latissimus dorsi (LD) transfer combined with reverse total shoulder arthroplasty (rTSA) to restore external rotation (ER) strength is clinically well-established, but studies directly comparing LD and intact teres minor (TM) strength are lacking. Also, variations in eccentricity in the glenoid components have been under-estimated. METHODS: We used eight fresh-frozen cadaveric shoulders in this study. LD transfer to the posterolateral aspect of the greater tuberosity was simulated using a cable pulley system. We explored six possible conditions: massive rotator cuff tears (RCTs) (supraspinatus and infraspinatus tears), massive RCTs with TM tears, massive RCTs with TM tears and LD transfer, and all three conditions complicated by centric and eccentric glenoid components. We measured the impingement-free range of motion (IFROM), the ER torque (N*m), the maximum abduction angle on successive loading of the middle deltoid (the abduction capacities), and the anterior dislocation forces. RESULTS: Use of an eccentric glenosphere was associated with greater IFROM in all directions, but the differences in ER strength, abduction capability, and anterior dislocation force compared with a centric design were not significant. LD transfer with TM tears showed significantly greater ER strength than massive RCTs condition at 30° (P<
0.05) and 60° (P<
0.05) of abduction. TM tears, with or without LD transfer, had less anterior stability than an intact TM (P<
0.05). Massive RCTs combined with TM tears tended to be associated with an increased abduction angle under the same deltoid load across all tested loads, revealing the key role played by the TM in joint stability and ER. CONCLUSIONS: LD transfer enhanced ER abduction strength in the absence of an intact TM, in contrast to a massive RCT with an intact TM. The TM was found to play a significant role in stability. Greater ROM in all directions was achieved with an eccentric than with a centric glenosphere.