Objective To analyze the effect of stump-preserving repair on rotator cuff healing and shoulder function for degenerative total rotator cuff tears. Methods A clinical data of 152 patients with degenerative total rotator cuff tears, who underwent arthroscopic repair between April 2019 and May 2022, was retrospectively analyzed. There were 76 males and 76 females with an average age of 55.4 years (range, 24-78 years). MRI was performed at 6 months postoperatively to evaluate the rotator cuff healing according to the Sugaya classification. Pre- and intra-operative related factors were included for univariate analysis, including age (≥60 years/<60 years), gender (male/female), passive activity disorder (yes/no), disease duration (≤3 months/>3 months), stump-preserving repair (yes/no), use of suture bridge technique (yes/no), shoulder joint abduction angle at knotting (<45°/≥45°), acromioplasty (yes/no), glucocorticoid injection (yes/no), time for patients to start postoperative passive exercise (≤2 weeks/>2 weeks), and time for patients to start postoperative active exercise (≤3 months/>3 months). The influencing factors of tendon healing were screened; further logistic regression was used to conduct multivariate analysis to screen for risk factors. Two sets of data were balanced by propensity score matching. The American Shoulder and Elbow Surgeons (ASES) score and Constant-Murley score of shoulder joint function at 6 and 12 months postoperatively, as well as rotator cuff healing rate at 6 months postoperatively, were compared between groups based on whether or not stump-preserving repair was used. Results All patients were followed up 12-33 months (mean, 23.8 months). MRI at 25-31 weeks postoperatively showed the 121 cases of rotator cuff healing and 31 cases of non healing. Univariate analysis showed that the disease duration, stump-preserving repair, shoulder joint abduction angle at knotting, and the time for patients to start postoperative active exercise were the influencing factors of rotator cuff healing (P<0.05). Multivariate analysis showed that non-stump-preserving repair, shoulder abduction angle more than 45° at knotting, and the time to start active exercise within 3 months postoperatively were risk factors affecting rotator cuff healing (P<0.05). A total of 51 pairs of cases were matched based on the grouping criteria of whether the disease duration exceeded 3 months, whether the shoulder abduction angle at knotting exceeded 45°, and whether the time to start postoperative active exercise exceeded 3 months. The rotator cuff healing rate, ASES score, and Constant-Murley score of the preserving repair group at 6 months postoperatively were superior to those of the non-preserving repair group, and the differences were significant (P<0.05). There was no significant difference in ASES score and Constant-Murley score between the two groups at 12 months postoperatively (P>0.05). Conclusion For degenerative total rotator cuff tears, the stump-preserving repair can shorten the healing time and promote the shoulder function recovery, but has no significant effect on shoulder function at 1 year postoperatively.