Background Due to the nature of rare autoimmune diseases, the risk factors for membranous nephropathy (MN) remain undetermined. We conducted an exposome-wide Mendelian randomization (EWMR) study to identify the causative diseases of MN. Methods We constructed genetic instruments for clinical outcomes in the FinnGen Project. To estimate the causal effect of multiple disease traits on MN, we performed two-sample Mendelian randomization analysis. First, we utilized the full set of instrumental variables (IV_full) to identify traits suggestive of a causal effect with statistical significance. We then determined conclusive causative traits that fulfill two criteria: (1) IV_full consisted of two or more instruments and (2) the effect remained significant after excluding instruments located in the MHC region (IV_nomhc). Results We included 451 traits satisfying the eligibility criteria for EWMR analysis. EWMR utilizing IV_full revealed 53 clinical outcomes with a suggestive causal effect on MN. Although the effect was attenuated when using IV_nomhc, celiac disease (OR_full: 1.49; P_full: 2.52×10^-5; OR_nomhc: 1.57; P_nomhc: 1.77×10^-6) and colorectal cancer (OR_full: 1.53; P_full: 5.44×10^-6; OR_nomhc: 1.53; P_nomhc: 5.44×10^-6) were identified as conclusive causal traits of MN with robust statistical significance. Conclusion We revealed causal exposures of MN at the exposome level, suggesting the importance of preventive strategies for MN in individuals with celiac disease or colorectal cancer.