To explore the long-term effectiveness of arthroscopic partial repair in treatment of massive irreparable rotator cuff tears from both the radiological and clinical perspectives. A retrospective analysis was conducted on the clinical data of 24 patients (25 sides) with massive irreparable rotator cuff tears who met the inclusion criteria between May 2006 and September 2014. Among them, there were 17 males (18 sides) and 7 females (7 sides) with an age range of 43-67 years (mean, 55.0 years). There were 23 cases of unilateral injury and 1 case of bilateral injuries. All patients were treated with the arthroscopic partial repair. The active range of motion of forward elevation and abduction, external rotation, and internal rotation, as well as the muscle strength for forward flexion and external rotation, were recorded before operation, at the first postoperative follow-up, and at last follow-up. The American Association of Shoulder and Elbow Surgeons (ASES) score, the University of California at Los Angeles (UCLA) shoulder scoring, and Constant score were used to evaluate shoulder joint function. And the visual analogue scale (VAS) score was used to evaluate shoulder joint pain. MRI examination was performed. The signal-to-noise quotient (SNQ) was measured above the anchor point near the footprint area (m area) and above the glenoid (g area) in the oblique coronal T2 fat suppression sequence. The atrophy of the supraspinatus muscle was evaluated using the tangent sign. The global fatty degeneration index (GFDI) was measured to assess fat infiltration in the supraspinatus muscle, infraspinatus muscle, teres minor muscle, upper and lower parts of the subscapularis muscle. The mean GFDI (GFDI-5) of 5 muscles was calculated. The incisions healed by first intention. All patients were followed up with the first follow-up time of 1.0-1.7 years (mean, 1.3 years) and the last follow-up time of 7-11 years (mean, 8.4 years). At last follow-up, the range of motion and muscle strength of forward elevation and abduction, ASES score, Constant score, UCLA score, and VAS score of the patients significantly improved when compared with those before operation ( P<0.05). Compared with the first follow-up, except for a significant increase in ASES score ( P<0.05), there was no significant difference in the other indicators ( P>0.05). Compared with those before operation, the degree of supraspinatus muscle infiltration worsened at last follow-up ( P<0.05), GFDI-5 increased significantly ( P<0.05), and there was significant difference in the tangent sign ( P<0.05); while there was no significant difference in the infiltration degree of infraspinatus muscle, teres minor muscle, and subscapularis muscle, upper and lower parts of the subscapularis muscle ( P>0.05). Compared with the first follow-up, the SNQm and SNQg decreased significantly at last follow-up ( P<0.05). At the first and last follow-up, there was no correlation between the SNQm and SNQg and the ASES score, Constant score, UCLA score, and VAS score of the shoulder ( P>0.05). Arthroscopic partial repair is effective in treating massive irreparable rotator cuff tear and significantly improves long-term shoulder joint function. For patients with severe preoperative fat infiltration involving a large number of tendons and poor quality of repairable tendons, it is suggested to consider other treatment methods.