BACKGROUND Recently, there have been a number of studies on the association between XRCC1 polymorphisms and childhood acute lymphoblastic leukemia (ALL) risk. However, the results of previous reports are inconsistent. Thus, we performed a meta-analysis to clarify the effects of XRCC1 variants on childhood ALL risk. METHODS A meta-analysis was performed to examine the association between XRCC1 polymorphisms (Arg399Gln, Arg194Trp, and Arg280His) and childhood ALL risk. We critically reviewed 7 studies with a total of 880 cases and 1311 controls for Arg399Gln polymorphism, 3 studies with a total of 345 cases and 554 controls for Arg280His polymorphism, and 6 studies with a total of 783 cases and 1180 controls for Arg194Trp polymorphism, respectively. Odds ratio (OR) and its 95% confidence interval (CI) were used. RESULTS Significant association between XRCC1 Arg399Gln polymorphism and childhood ALL risk was observed in total population analyses (OR(additive model) = 1.501, 95% CI 1.112-2.026, P(OR) = 0.008; OR(dominant model) = 1.316, 95% CI = 1.104-1.569, P(OR) = 0.002) and Asian subgroup analyses (OR(additive model) = 2.338, 95%CI = 1.254-4.359, P(OR) = 0.008; OR(dominant model) = 2.108, 95%CI = 1.498-2.967, P(OR) = 0.000). No association was detected in Caucasians, Metizo and mixed populations. Ethnicity was considered as a significant source of heterogeneity in the meta-regression model. For the other two XRCC1 polymorphisms, no association with childhood ALL risk was found. CONCLUSIONS The meta-analysis results suggested that XRCC1 Arg399Gln polymorphism might be associated with elevated childhood ALL risk among Asian population.