Range of motion after total shoulder arthroplasty is better than after reverse shoulder arthroplasty, however with similar clinical outcome. It is unclear if this difference can only be found in the different range of motion or also in the force generating capacity.(1) are isokinetically produced joint torques of reverse shoulder arthroplasty comparable to those of total shoulder arthroplasty? (2) Does this force-generating capacity correlate with functional outcome?Eighteen reverse shoulder arthroplasty patients (71years (SD 9years)) (21 shoulders, follow-up of 21months (SD 10months)) were recruited, 12 total shoulder arthroplasty patients (69years (SD 9years)) (14 shoulders, follow-up of 35months (SD 11months)). Pre- and post-operative Constant-Murley scores were obtained; two isokinetic protocols (ab-/adduction and ex-/internal rotations) at 60°/s were performed.Twelve of 18 reverse shoulder arthroplasty patients generated enough speed to perform the test (13 shoulders). Mean ab-/adduction torques are 16.3Nm (SD 5.6Nm) and 20.4Nm (SD 11.8Nm). All total shoulder arthroplasty patients generated enough speed (14 shoulders). Mean ab-/adduction torques are 32.1Nm (SD 13.3Nm) and 43.1Nm (SD 21.5Nm). Only 8 reverse shoulder arthroplasty patients (9 shoulders) could perform ex-/internal rotation tasks and all total shoulder arthroplasty patients. Mean ex-/internal rotation torques are 9.3Nm (SD 4.7Nm) and 9.2Nm (SD 2.1Nm) for reverse shoulder arthroplasty, and 17.9Nm (SD 7.7Nm) and 23.5Nm (SD 10.6Nm) for total shoulder arthroplasty. Significant correlations between sub-scores: activity, mobility and strength and external rotation torques for reverse shoulder arthroplasty. Moderate to strong correlation for sub-scores: strength in relation to abduction, adduction and internal rotation torques for total shoulder arthroplasty.Shoulders with a total shoulder arthroplasty are stronger. This can be explained by the absence of rotator cuff muscles and (probably) medialized center of rotation in reverse shoulder arthroplasty. The strong correlation between external rotation torques and post-operative Constant-Murley sub-scores demonstrates that external rotation is essential for good clinical functioning in reverse shoulder arthroplasty.