Current research on human rotator cuff pathology relies on superficial biopsy specimens. It is unclear whether these biopsies are representative of overall muscle quality. The purpose of this study is to use magnetic resonance imaging with iterative decomposition of echoes of asymmetric length sequencing to investigate variability of fatty infiltration within the supraspinatus and infraspinatus muscle. We retrospectively identified 45 patients who underwent arthroscopic rotator cuff repair with preoperative iterative decomposition of echoes of asymmetric length imaging completed. The supraspinatus and infraspinatus were segmented on 4 consecutive slices, including the scapular Y, 2 slices medial, and 1 slice lateral. Intramuscular fat was measured in multiple regions for both supraspinatus (whole muscle, anterior, posterior, superficial band, anterior band, and posterior band) and infraspinatus (whole muscle, superior, inferior, superficial band, superior band, and inferior band). Comparisons of intramuscular fat were determined with Wilcoxon sign-rank tests. Analysis of variance was used to compare between the 4 consecutive slices. Significance was defined as P < .05. Magnetic resonance imaging showed 31 full-thickness supraspinatus tears, 10 partial-thickness supraspinatus tears, and 4 intact supraspinatus tendons and 3 full-thickness infraspinatus tears, 2 partial-thickness infraspinatus tears, and 40 intact infraspinatus tendons. The anterior supraspinatus contained significantly higher fat content than the posterior supraspinatus (7.4% ± 7.4% vs. 5.4% ± 5.7%, P = .003). The superior and inferior halves of the infraspinatus were not different from each other (P = .11). The superficial band did not differ from the whole muscle in both supraspinatus (P = .14) and infraspinatus (P = .20). However, the anterior band of the supraspinatus had significantly more fat than the posterior band (8.2% ± 9.3% vs. 5.0% ± 5.7%, respectively, P < .0001), and the superior band of the infraspinatus had significantly more fat than the inferior band (5.2% ± 4.8% vs. 4.2% ± 5.3%, respectively, P = .03). There was no difference between all 4 medial and lateral slices in the supraspinatus (P = .92) and infraspinatus (P = .90). Fat fractions within the supraspinatus and infraspinatus demonstrate significant spatial variability that may influence interpretation of local biopsy samples. Future biopsy studies may benefit from multiple samples between different specific locations.