With exercise being more frequently utilized in treatment for obstructive sleep apnea (OSA), a systematic review of the intervention efficacy of exercise on OSA is necessary. PubMed, EBSCO, Web of Science, VIP, and CNKI databases were searched to collect randomized controlled trials (RCTs) of exercise applied to OSA from January 2000 to January 2022. The literature screening, data extraction, and risk of bias assessment of included studies were conducted independently by two reviewers. Meta-analysis was then performed using Rev Man 5.4 software. A total of 9 RCTs were included, including 444 patients. Compared with the control group, exercise made an improvement in apnea–hypopnea index (AHI) [MD = −6.65, 95% CI (−7.77, −5.53), p < 0.00001], minimum oxygen saturation (SaO2min%) [MD = 1.67, 95% CI (0.82, 2.52), p = 0.0001], peak oxygen uptake (VO2peak) [SMD = 0.54, 95% CI (0.31, 0.78), p < 0.00001], Pittsburgh sleep quality index (PSQI) [MD = −2.08, 95% CI (−3.95, −0.21), p = 0.03], and Epworth Sleepiness Scale (ESS) values [MD = −1.64, 95% CI, (−3.07, −0.22), p = 0.02]. However, there were no significant changes in body mass index (BMI). As for the results of subgroup analysis, aerobic exercise combined with resistance exercise [MD = −7.36, 95% CI (−8.64, −6.08), p < 0.00001] had a better effect on AHI reduction than aerobic exercise alone [MD = −4.36, 95% CI (−6.67, −2.06), p = 0.0002]. This systematic review demonstrates that exercise reduces the severity of OSA with no changes in BMI, and the effect of aerobic exercise combined with resistance training is better than aerobic exercise alone in AHI reduction. Exercise also improves cardiopulmonary fitness, sleep quality, and excessive daytime sleepiness.