Background: Epidemiologic studies have yielded inconsistent findings between breastfeeding and epithelial ovarian cancer (EOC) risk. Objective: We performed a meta-analysis to summarize available evidence of the association between breastfeeding and breastfeeding duration and EOC risk from published cohort and case-control studies. Design: Relevant published studies were identified by a search of MEDLINE through December 2012. Two authors (T-TG and Q-JW) independently performed the eligibility evaluation and data abstraction. Study-specific RRs from individual studies were pooled by using a random-effects model, and heterogeneity and publication-bias analyses were conducted. Results: Five prospective and 30 case-control studies were included in this analysis. The pooled RR for ever compared with never breastfeeding was 0.76 (95% CI: 0.69, 0.83), with moderate heterogeneity (Q = 69.4, P < 0.001, I(2) = 55.3%). Risk of EOC decreased by 8% for every 5-mo increase in the duration of breastfeeding (RR: 0.92; 95% CI: 0.90, 0.95). The risk reduction was similar for borderline and invasive EOC and was consistent within case-control and cohort studies. Conclusions: Results of this meta-analysis support the hypothesis that ever breastfeeding and a longer duration of breastfeeding are associated with lower risks of EOC. Additional research is warranted to focus on the association with cancer grade and histologic subtypes of EOC.