The proximal humerus, a common site for osteoporotic fractures, is frequently overlooked in osteoporosis evaluations. This study aimed to evaluate the relationship between the conventional bone mineral density (BMD) measurement (at the lumbar spine and femur) and the BMD measurement at both proximal humeri (the asymptomatic side and the side with a rotator cuff tear [RCT]) in patients with unilateral RCT. Furthermore, we investigated clinical features indicative of osteoporosis in RCT patients and assessed the utility of proximal humerus BMD measurements. From April 2020 to September 2020, 87 patients who underwent arthroscopic repairs for unilateral RCTs were examined for age, onset, body mass index, menopause duration, passive range of motion, global fatty degeneration index, and RCT and retraction size. The regions of interest (ROIs) for the conventional BMD included the lumbar spine, femur neck, femur trochanter, and total femur. For the proximal humerus BMD, the ROIs included the head, lesser tuberosity, greater tuberosity (medial, middle, and lateral rows), and total humerus. The conventional BMD of the lumbar spine, femur neck, femur trochanter, and femur total were 1.090, 0.856, 0.781, and 0.945 g/cm2, respectively. The head, lesser tuberosity, greater tuberosity (medial, middle, and lateral rows), and total BMD of the asymptomatic-side proximal humerus were 0.547, 0.544, 0.697, 0.642, 0.554, and 0.610 g/cm2, respectively. The average of proximal humerus BMD was significantly lower than that of conventional BMD (p < 0.001). All ROIs BMD of the RCT-side proximal humerus were 0.497, 0.507, 0.619, 0.598, 0.517, and 0.560 g/cm2. There was no correlation between the conventional BMD and each proximal humerus BMD. All ROI BMD of the RCT-side proximal humerus was not significant in the multiple regression analysis with age, onset, body mass index, passive range of motion, global fatty degeneration index, and RCT and retraction size (p > 0.05). The proximal humerus BMD showed a completely different trend from that of conventional BMD and had no significant association with clinical features. Therefore, the proximal humerus BMD needs to be measured separately from the conventional BMD, as it may provide important information before rotator cuff repair surgery.