Background The etiology of bladder cancer is not yet well known. In this study, we want to evaluate the effect of polymorphisms of genes that have an epigenetic effect (MTHFR, DNMT3A/B) on the susceptibility to develop bladder cancer (BC). Methods A systematic review was performed for MTHFR, DNMT3A, and DNMT3B, followed by a meta-analysis conducted for rs1801131, rs1801133, rs2274976, rs1550117, and rs1569686 SNPs. A sensitivity and a subgroup analysis were then used. Results 20 studies were included, where no statistically significant association between any of the analyzed SNPs and the occurrence of BC was detected. Subgroup analysis revealed a statistically significant association in North African population with rs1801133: TT vs. TC + CC (P = 0.013; OR 95% CI = 0.52 [0.311–0.872]); TT vs.TC (P = 0.003; OR 95% CI = 0.448 [0.261–0.769]) and in North American population with rs1801131: CC vs. CA (P = 0.039; OR 95% CI = 0.71 [0.523–0.984]). A sensitivity analysis revealed that there is a statistically significant association between rs1801131 and the occurrence of BC (OR = 0.79, 95%CI [0.65–0.97]), (OR = 0.80, 95%CI [0.65–0.98]) and (OR = 0.78, 95%CI [0.63–0.96]) which correspond to CC vs. CA + AA; CC vs. CA; and CC vs. AA genetic models. Conclusion This is the first study to assess the effect of DNMTs on bladder cancer risk. No statistically significant association was found between polymorphisms of MTHFR, DNMT3A/B genes and bladder cancer development, except for the North African and the North American populations with rs1801133 and  rs1801131, respectively, with a protective effect of rs1801131 based on a sensitivity analysis.