Cutibacterium acnes (C. acnes), a common pathogen, contributes significantly to infections in shoulder surgery. Prevention of shoulder infection is crucial to improve postoperative functional recovery and reduce costs. This study aimed to perform a systematic review and meta-analysis to assess the safety and efficacy of 5% benzoyl peroxide (BPO) application in the shoulder to decrease C. acnes. Three electronic databases were searched as follows: PubMed, Embase, and the Cochrane Library databases. Data extraction for this study was performed by two independent reviewers, and only level I and level II studies were included. The outcome data sources of individual studies were pooled. The fixed-effect model was used to determine the meta-analysis. There were five level I studies and five level II studies. The results showed that the 5% BPO group had a lower risk of C. acnes positivity [OR, 0.21 (0.15, 0.30), I 2 = 24, p < 0.00001]. The pooled analysis results showed that there was no significant difference in the ability of 5% BPO and 5% BPO + clindamycin to reduce C. acnes. However, the lower rate of adverse events was significantly in favour of the non-BPO group compared with the 5% BPO group. BPO can decrease C. acnes in the shoulder to prevent infection. However, the combination of BPO and clindamycin does not enhance this effect further. II, Systematic review and meta-analysis.