Background: Hyperoxia during anesthesia can increase cellular oxidative stress, and perioperative antioxidant treatment may reduce the resulting damage. The aim of this review was to evaluate risks and benefits of antioxidant treatment in surgical patients. We hypothesized that antioxidant treatment reduced mortality compared to placebo/no intervention. Methods: This systematic review with meta-analyses and trial sequential analysis (TSA) was conducted using Cochrane standards and GRADE methodology. Randomized clinical trials comparing perioperative antioxidant treatment vs. placebo/no intervention in adults were included. Primary outcome was mortality at longest follow-up. Results: Ninety-seven trials with 8156 patients were included. The most common interventions were N-Acetylcysteine (36 trials) and vitamin C (29 trials). Trials were primarily performed during cardiac surgery (53 trials). Fifty-six trials with 4890 patients reported information on mortality (243 events). The meta-analysis of mortality at longest follow-up showed a reduced mortality in antioxidant treated patients (RR 0.74, 95% CI 0.59; 0.94, I² 0%), however, TSA-adjusted CI was broadened (0.55-1.02) and only 31% of the required information size was reached. Furthermore, in the subgroup of three trials with overall low risk of bias the RR for mortality was 1.18 (95% CI 0.39, 3.63). Based on GRADE, our findings are of low quality of evidence due to high risk of bias, imprecision, and indirectness. Conclusion: We found a 26% relative risk reduction of mortality in surgical patients treated with antioxidants but the quality of evidence supporting our findings is low and influenced by clinical heterogeneity and high risk of systematic- and random errors.