BACKGROUND The degree of fatty infiltration of the rotator cuff muscle is typically evaluated using the Goutallier-Fuchs grading system, but its consistency remains controversial. This study aimed to evaluate a new quantified measurement of fatty infiltration based on three-dimensionally reconstructed volumetric data obtained from magnetic resonance images of non-pathologic shoulders using open-source software. METHODS Fourteen shoulder 3-T magnetic resonance images (8 men, 6 women) without lesions obtained between 2010 and 2017 were analysed. Slicer version 4.6.2 was used to semi-automatically reconstruct the three-dimensional volumetric data from T2 sagittal oblique images and to differentiate fat tissue from rotator cuff muscle using the difference in signal intensity. RESULTS The cutoff value for dividing muscle and fat was 508.9. The inter-class and intra-class correlations of each rotator cuff muscle and fat tissue were >0.9 (all P < 0.001). The mean muscle volume of the supraspinatus, infraspinatus, teres minor, and subscapularis were 15.2, 20.9, 13.3, and 29.7 mL, respectively. The muscle volume of the men was greater than that of the women (all P < 0.001), and the fat infiltration ratio was positively correlated with body mass index (all P < 0.05). CONCLUSIONS The semi-automated quantified measurement of fatty infiltration of the rotator cuff muscles using magnetic resonance imaging and Slicer software presented excellent consistency. This technique could be an alternative measurement to complement the weak consistency of the Goutallier-Fuchs grading system. However, to reduce the error of measurement, this study evaluated non-pathologic shoulders. Therefore, further study using magnetic resonance imaging of pathologic shoulders is necessary for actual clinical application. METHODS Level IV, case series, diagnostic study.