The purpose of this study was to characterize the prevalence of taper damage in retrieved modular shoulder arthroplasty components. We matched this cohort with historical and contemporary total hip replacements. The components were matched primarily for implantation time, but also for alloy combination (mixed versus similar) and patient age (as a proxy for patient activity), when available. Modular components were evaluated for mechanically assisted crevice corrosion using a semiquantitative four-point scoring system. In total, 92 anatomical humeral heads and 67 associated stems with an average implantation time of 4.6 ± 4.3 years were analyzed. Additionally, 79 femoral heads and 61 associated stems with an average implantation time of 4.5 ± 4.4 years were matched to shoulder components and analyzed. Design and patient factors were assessed as potential predictors of taper damage. Evidence of damage was observed on 46 % of humeral heads and 66 % of the humeral stems. In the hips, 47 % of femoral heads and 44 % of femoral stems had evidence of damage. Mixed alloy pairs (p < 0.0001; stem side only), taper rigidity (p < 0.0001), and implantation time (p = 0.01) were associated with taper corrosion. The results from this study supported the prevalence of taper damage that is similar between anatomical shoulder and conventional hip arthroplasty. The clinical implications of these findings have not yet been determined.