Objective:To provide a more robust assessment to the effect of cytochrome P2D6 (CYP2D6) polymorphisms on the risk of acute leukemia (AL), and to evaluate the association between the two most commonly studied CYP2D6 polymorphisms (CYP2D6*3 and CYP2D6*4) and AL risk by meta-analysis. Methods:All case-control studies investigating an association between the CYP2D6*3 or CYP2D6*4 polymorphisms and AL risk were included. Either fixed-effects or random-effects models were applied to combine odds ratios (ORs) and 95%conifdence intervals (CIs) by RevMan 5.1. Q-statistic was used to evaluate the heterogeneity, and both Egger’s test and funnel plots were used to assess publication bias. Results:Six studies were included in the meta-analysis. The results we acquired were that the OR value and 95%CI of CYP2D6*4 wild type, heterozygous mutant and homozygous mutant were 0.94 (0.66-1.35), 1.04(0.74-1.45) and 1.63 (0.95-2.81), respectively with Z=0.33, 0.23 and 1.76 (P>0.05), indicating that there was no signiifcant association between CYP2D6*4 polymorphism and the risk of AL. We also performed subgroup analysis by the AL immunophenotype for those groups with heterogeneity. The results of the combined analysis of CYP2D6*4 wild type, heterozygous mutant and acute lymphoblastic leukemia (ALL) were Z=0.08, 0.08 (P>0.05), for acute myeloblastic leukemia (AML) were Z=0.17, 0.26 (P>0.05), indicating that there was no signiifcant association between CYP2D6*4 polymorphism and the development of both ALL and AML. Conclusion:CYP2D6 polymorphisms are not associated with AL risk.