Objective To compare the effect of whether rotator cuff pathological tissue cleaning and tendon insertion reconstruction or not under shoulder arthroscopy on the surgical treatment of rotator cuff tendinopathy. Methods From September 2015 to January 2017, the clinical data of forty-one patients with rotator cuff tendinopathy treated by surgical medicine ward in Shengjing Hospital of China Medical University were retrospectively analyzed.According to different surgical methods, 41 patients were divided into A group and B group.In group A, 21 patients were treated with shoulder arthroscopic bursa debridement combined with selective acromioplasty.In group B, 20 patients were treated with cleanup of pathological changes tissue and reconstruction of tendon insertion on the basis of A. All patients were followed up for one year.Constant-Murley shoulder function score, visual analog scale (VAS), University of California Los Angeles (UCLA) and Hawkins sign, arc of pain sign and infraspinatus muscle test were used as the evaluation indexes.The shoulder joint function was evaluated so as to compare the efficacy of two surgical methods for rotator cuff disease. Results Compared with the preoperative, the postoperative VAS score, UCLA score and Constant-Murley score in both groups were significantly improved.In group A, the preoperative in group A (5.0±1.3)points, (15.5±1.4) points, (65.1±5.7) points) were increased to (0.7±0.4) points, (33.0±1.7) points, (90.9±3.1) points.The preoperative scores of group B were (5.2±1.0)points, (15.6±2.0) points, (65.4±5.9) points, and increased to (0.4±0.5) points, (34.3±0.9) points, (93.3±2.3) points respectively, and the differences were statistically significant (group A: t=14.77, P<0.001; t=- 74.44, P<0.001; t=- 29.19, P<0.001; group B: t=23.13, P<0.001; t=- 61.52, P<0.001; t=- 26.38, P<0.001). The UCLA score and Constant-Murley score of shoulder joint in group B were higher than those in group A (t=-3.27, P=0.003; t=-2.90, P=0.007). Postoperative positive rates of Hawkins sign, pain arc sign and infraspinatus muscle test in group A decreased from 90% (18/20), 90% (18/20), 95% (19/20) to 15% (3/20), 10% (2/20) and 15% (3/20) respectively.There was a significant difference between preoperative and postoperative (χ2=13.067, P<0.001). χ2=14.063, P<0.001; χ2=14.063, P<0.001), group B decreased from 90.5% (19/21), 85.6% (18/21), 90.5% (19/21) to 9.5% (2/21), 4.8%(1/21), 9.5% (2/21). There were significant differences between preoperative and postoperative (χ2=15.059, P<0.001; χ2=12.500, P<0.001; χ2=15.059, P<0.001). Conclusion Both group A and group B are effective in the treatment of rotator cuff tendon disease, and group B is more effective than group A in the treatment of rotator cuff tendon disease. Key words: Shoulder joint; Rotator cuff tendinopathy; Arthroscope; Clinical efficacy