Previous studies concerning the association between nonoccupational physical activity (PA) and risk of ovarian cancer yielded mixed results. We investigated the association by performing a meta-analysis. Relevant studies were identified by searching PubMed and EMBASE to June 2014. We calculated the summary relative risks (RRs) and 95% confidence intervals (CIs) using random-effects models. The dose-response relationship was assessed by restricted cubic spline model and multivariate random-effect meta-regression. Nine cohort studies and ten case-control studies involving 730,703 participants and 9,459 cases of ovarian cancer were selected for meta-analysis. The analyses showed that individuals who participated in any amount of nonoccupational PA had a RR of 0.92 (95% CI = 0.84-1.00) for risk of ovarian cancer. Those who participated in high or moderate nonoccupational PA had a RR of ovarian cancer risk of 0.89 (95% CI = 0.79-1.01) and 0.91 (95% CI = 0.85-0.99), respectively. Stratifying by study design and cancer subtype (borderline and invasive tumors), inverse association was only found in case-control studies. A linear but not significant dose-response relationship was found between nonoccupational PA and ovarian cancer risk. In conclusion, a weak inverse association exists between nonoccupational PA and the risk of ovarian cancer. Regarding the significant heterogeneity among included studies, confirmation in further prospective cohort studies with more accurate assessment of PA level is essential.