In spite of widespread application around the world, there has been controversy on the cerebral and cardiac protection efficacy of carbon dioxide insufflation (CDI) during open heart surgery. To make a comprehensive evaluation, we screened all relevant published randomized controlled trials to perform the first systematic review and meta-analysis for CDI during open heart surgery. We searched PubMed, EMBASE, the Cochrane Controlled Clinical Trial register, WANFAN, CQVIP, and CNKI database for published articles. Randomized controlled trials were included when the research provided data of neurological complications postoperatively, creatinine kinase, MB isoenzyme (CK-MB) on the first postoperative day, or all-cause mortality. We chose a fixed-effects model when the trials showed low heterogeneity, otherwise a random effects model was used. The quality of studies was assessed by modified Jadad scale. Four studies were included in this meta-analysis. The overall pooled relative risk (RR) for neurological complications was 1.59, 95% confidence interval (CI) = [0.57, 4.46], and the z-score for overall effect was 0.89 (P = 0.37). The standardized mean difference of the CK-MB between groups was 1.15, 95% CI = [-1.27, 3.56], and the z-score for overall effect was 0.93 (P = 0.35). The overall pooled RR for all-cause mortality was 0.5, 95% CI = [0.16, 1.64], and the z-score for overall effect was 1.14 (P = 0.25). There was no significant difference between groups. Because of the insufficiency of powerful evidences, the cerebral and cardiac protection efficacy of CDI during open heart surgery needs to be further verified by more high-quality trials.