Current research on the relationship between environmental endocrine disrupting chemicals (EDCs) and breast cancer remains insufficient, with limited evidence and inconsistent conclusions. Mendelian randomization (MR) is a robust method for establishing causality, as it reduces biases from confounding factors and reverse causation. This study uses MR to investigate the effects of three types of EDCs, including bisphenols, parabens, and phthalates, on the risk of overall breast cancer and its subtypes-Luminal A, Luminal B, triple negative, human epidermal growth factor receptor 2-enriched, and estrogen receptor-positive/negative. The study also examines the 1400 blood metabolome as potential mediators and explores EDCs-associated DNA methylation changes as potential factors, with a focus on European populations. Our results shows that n-butyl paraben (n-BuP) is positively associated with Luminal A, mono-methyl phthalate is negatively associated with Luminal B, and mono-iso-butyl phthalate (MiBP) is positively associated with triple negative breast cancer (TNBC). Mediation analysis reveals that blood metabolites, such as caffeic acid sulfate and the caffeine-to-paraxanthine ratio, mediate the effect of n-BuP on Luminal A, while methylsuccinate mediate the effect of MiBP on TNBC. Epigenetic analysis shows associations between EDCs exposure-related DNA methylation changes at specific CpG sites (cg26325335, cg08537847, cg27454300) and different breast cancer risks. These findings not only suggest potential biomarkers for early detection and intervention but also underscore the imperative for further research to rigorously validate these associations.