Abstract:To evaluate the effects of arthroscopic debridement (DB), partial repair (PR) and complete repair (CR) on massive rotator cuff tears (MRCT). Methods 68 patients with shoulder joint tear were included, including at least 3 tendons mRCT, and treated with arthroscopic DB (n = 22), PR (n = 22) and CR (n = 23), respectively. Functional and subjective parameters were used to evaluate the treatment effect before and 45 months after surgery in all patients, and MRI was used to evaluate the form of preoperative tendon rupture and postoperative repair integrity. Results There was no significant difference in preoperative parameters of the patients except fatty degeneration. Compared with preoperative parameters, the postoperative indicators of the patients in the DB, PR and CR groups were significantly improved (P < 0.05). However, post-operative comparisons revealed similar scores with DB [constant score (65.8 ± 14.7), DASH (24.1 ± 20.6)] and PR [CS (67.5 ± 9.9), P > 0.05; DASH (20.5 ± 14.4), P > 0.05], while CR were significantly better [CS(80.3 ± 8.9); DASH (7.0 ± 8.7); all P < 0.05]. Force couple restoration of PR did not significantly influence outcome. Re-tear rates with CR (29.0 %) were lower compared to PR (53.0%). Intact CR compared to intact PR showed better [CS (83.4 ± 7.3) vs. (68.5 ± 10.6), P = 0.009] and DASH [(5.4 ± 8.3) vs. (21.2 ± 9.5), P = 0.006]. The vast majority of patients were satisfied with their arthroscopic procedure (group DB 87.0%, PR 86.0%, CR 91.0%). Conclusion Arthroscopic DB, PR, and CR were effective in treating MRCT involving at least three tendons. Reparability of MRCT was influenced by fatty degeneration of the muscles. However, CR showed the most favourable short-term improvements.