Currently, autografts and allografts are largely used to treat large or massive rotator cuff tear (RCT), without any evidence in favour of one graft or the other. The purpose of this study was to determine the rate of retear of autograft and allograft in the treatment of large or massive posterosuperior RCT. Method The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were followed to perform this systematic review and meta-analysis of the results in the literature as well as the presentation of results. A search of the literature was performed in the electronic databases MEDLINE, Scopus, Embase, and the Cochrane Library. The quality of the included studies was evaluated according to the MINORS (Methodological Index for Nonrandomized Studies) score. Inclusion criteria were studies in English evaluating clinical and radiological results of surgical treatment with autograft or allograft for large or massive RCT since 2008. The main criterion was the retear rate of the graft assessed on MRI or US scan at 1-year minimum follow-up. Partial tear were classified as “tear”. Results The overall retear rate was 23.6% (15.5–32.7) at a mean follow-up of 18.4 ± 7.8 (12–36) months. There was no significant difference between the two kinds of graft, with a retear rate of 27.0% (15.4–40.2) and 20.9% (9.9–34.2) with autograft and allograft respectively (n.s.). Similar improvements of functional scores (+ 28.8 to 38.4 points for the Constant score, + 33.6 to 38.4 points for the ASES, and − 4.0 to − 4.1 points for pain-VAS) were reported in the two groups after at 27.2 ± 11.1 (12–48) months. The rate of complications except retear was 1.8% (0.2, 3.7) with autograft and 0.5% (0.8, 1.8) with allograft (n.s.). Conclusion The use of autograft and allograft for the treatment of large or massive RCT leads to similar retear rate and clinical outcomes at short to medium terms. Level of evidence Level IV