HYPOTHESIS: Anterior instability is one of the most common complications in reverse shoulder replacement. This study hypothesized that intrinsic stability of a reverse prosthesis varies with the degree of version of the humerus and glenoid components. This should provide guidelines on how to best position the implant components to decrease the rate of dislocation. MATERIALS AND METHODS: Resistance to anterior dislocation of a reverse implant was measured in a mechanical testing machine by means of the stability ratio (ratio of peak dislocation/axial compressive forces). Versions of the humeral and glenoid components were modified in 10 degrees steps in the 90 degrees abducted and resting positions. RESULTS: In both tested positions, the effect of humeral component version was highly significant. Only a glenoid component retroversion of 20 degrees led to a statistically significant drop in stability ratio for the 20 degrees abducted position. Intrinsic stability in the typical component positioning (neutral glenoid version and 20 degrees humeral retroversion) yielded comparably low intrinsic stability, which could only be improved by increasing anteversion of the humeral component. DISCUSSION: Version of the humeral component is the critical factor for intrinsic stability. Version of the glenoid component is less important for intrinsic stability, but special care should be given to avoid retroversions of more than 10 degrees . Within this range, the surgeon can concentrate primarily on other parameters critical for long-term outcome (range of motion, secure fixation) when choosing the appropriate glenoid version. CONCLUSION: Anterior stability can be improved by implanting the humeral component in neutral or with some anteversion.