Estrogen signal medicated by estrogen receptor (ER), which is involved in various diseases related to steroid hormone, such as cancer. A number of association studies have focused on ESR2 polymorphisms to investigate the relationship with cancer risk. However, the results are inconsistent and inconclusive. To examine this controversy, 33 studies were enrolled for the pooled analysis for three poly-morphisms (rs3020450, rs4986938, and rs1256049) in cancer risk using odds ratios (ORs) with 95% confidence intervals (CIs). Regarding rs4986938, A allele was associated with decreased breast cancer. Ethnicity subgroup analysis observed a decreased risk in both Asian and Caucasian descendent. Regarding rs1256049, cancer type subgroup analysis revealed a significant association with increased prostate and endometrial cancer risk. rs3020450 was not associated with cancer risk in any model. Further studies for clarifying the roles of ESR2 polymorphisms in cancer risk seem of vital importance.