Background <p>To explore the causal association between genetically predicted Epstein–Barr virus (EBV) infection and breast, cervical, endometrial, and ovarian cancers.</p> Methods <p>Public data from genome-wide association studies of EBV infection and cancers were used to perform a two-sample Mendelian randomization (MR) study. The inverse variance weighted (IVW) method was the main analysis method, while the weighted median, weighted mode, and MR-Egger methods were used to test the robustness of the IVW results. Cochran’s Q-test, MR-Egger regression, MR-PRESSO, and leave-one-out methods were used to detect heterogeneity, horizontal pleiotropy, outliers, and SNPs driving the results, respectively.</p> Results <p>There were no statistically significant causal associations between genetically predicted EBV infection and genetically predicted endometrial (OR = 1.55, 95%CI: 0.71–3.37, <i>P</i> = 0.27), cervical (OR = 0.96, 95%CI: 0.42–2.22, <i>P</i> = 0.93), ovarian (OR = 1.55, 95%CI: 0.72–3.33, <i>P</i> = 0.26), or breast (OR = 1.02, 95%CI: 0.79–1.33, <i>P</i> = 0.85) cancers. The raw MR-PRESSO analyses support the IVW results but detected one outlier for the endometrial cancer analysis. After removing the outlier, a suggestive association was observed between EBV infection and endometrial cancer (OR = 2.43, 95%CI: 1.19–4.95, <i>P</i> = 0.024). Cochran’s Q-test detected heterogeneity for endometrial cancer (<i>P</i> = 0.04). The MR-Egger regression analysis revealed no horizontal pleiotropy. The leave-one-out analysis revealed no SNPs driving the results by themselves.</p> Conclusions <p>This MR study showed a possible causal association between EBV infection and the risk of endometrial cancer. This association warrants additional study.</p>