Objective: Exercise is reported to be beneficial for breast cancer. However, the results seem inconsistent. We conducted this systematic review and meta-analysis of animal experimental studies to fully understand the effect of exercise on breast cancer in animal model.Methods: We searched databases from inception to April 2022 and manually searched related references to retrieve eligible studies. We screened eligible studies and extracted related data. We assessed the risk of bias and reporting quality using the SYstematic Review Centre for Laboratory animal Experimentation Risk of Bias tool and the Animal Research: Reporting of In Vivo Experiments guidelines 2.0, respectively. We summarized the study characteristics and findings of included studies and conducted meta-analysis with RevMan software. Subgroup analysis and sensitivity analysis were also performed.Results: We identified 537 potential literatures and included 47 articles for analysis. According to the results of risk of bias assessment, only selective outcome reporting was in low risk of bias. Items of sequence generation, random outcome assessment, and incomplete outcome data were rated as high risk of bias. Most of other items were rated unclear risk of bias. In reporting quality assessment, all included articles reported grouping method and experimental procedures. However, no study provided information of the study protocol registration. Meta-analysis showed that, compared with sedentary lifestyle, exercise reduced more tumor weight (MD = −0.76, 95%CI −0.88 to −0.63, p = 0.85, I2 = 0%) and tumor number per animal (MD = −0.61, 95%CI −0.91 to −0.31, p = 0.34, I2 = 8%). Exercise decreased more tumor incidence than sedentary lifestyle both in motorized wheel/high-intensity (OR = 0.22, 95%CI 0.11 to 0.46, p = 0.09, I2 = 41%) and free wheel/low-intensity treadmill running (OR = 0.45, 95%CI 0.14 to 1.44, p = 0.04, I2 = 60%). Sensitivity analysis showed that the results were robust.Conclusion: Exercise could reduce tumor weight, number of tumors per animal, and incidence of tumor in breast cancer model of mice and rats. However, the risk of bias items and reporting guidelines in preclinical studies should be concerned. Future research should consider standards of conducting and reporting preclinical studies and choose suitable exercise protocol for higher quality evidence of exercise for breast cancer.