BACKGROUND : The study objective was to compare the reaming congruency of hand, power, and visual feedback axial alignment-guided (Marksman) reamers on glenoid models. We hypothesized that (1) a significant difference in average reaming deviation would be found between reamer types and (2) less ream-to-ream variation would occur with the Marksman reamer. METHODS : Retroversion correction of 18 identical Sawbones glenoid models was conducted using either a hand, power, or Marksman reamer with a 40-mm curvature radius. Glenoid correction with either 0° or 10° reaming was conducted in triplicate sets for each reamer. Reamed glenoid computed tomography images were 3-dimensionally reconstructed using Mimics Medical software (version 21.0). Congruency between the glenoid surface and a 3-dimensional sphere with a 40-mm curvature radius was analyzed. Average deviation and ream-to-ream variance were compared between the hand, power, and Marksman reamer groups. RESULTS : The power reamer demonstrated the smallest median deviation (0.08 mm; interquartile range [IQR], 0.07-0.19 mm), followed by the Marksman (0.09 mm; IQR, 0.08-0.17 mm) and hand (0.11 mm; IQR, 0.10-0.13 mm) reamers. Kruskal-Wallis analysis indicated no significant difference in deviation among the 3 reaming methods (P = .42). The Marksman reamer demonstrated the least variance (0.0034 mm), followed by the power (0.0076 mm) and hand (0.0093 mm) reamers. The results of the Conover squared ranks test indicated no significant difference in variance among the 3 reaming methods (P = .32). CONCLUSIONS : Our findings showed no statistically significant difference in the accuracy or consistency of reaming between reamer types. Trends showed less variance in the Marksman reamer group compared with the hand and power reamer groups, although differences in variation between groups were not statistically significant.