Previous studies have proposed the association between catechol-o-methyltransferase (COMT) Val158Met polymorphism and the risk of Alzheimer's disease (AD). However, there has been no solid conclusion from those reports because of the limited sample size and/or racial diversity. Here we aimed to give an in-depth and accurate evaluation of this association by performing a comprehensive literature search on websites including PubMed, EMBASE, Alzgene database, Cochrane library and Ovid database. In further meta-analysis we identified 10 case–control studies which contained total 2777 cases and 2829 controls. Summarized odds ratios (ORs) and corresponding 95% confidence intervals (CIs) were assessed to reveal the association. There was no significant association between COMT Val158Met polymorphism and AD susceptibility in all genetic models for the entire population. In the contrast the subgroup analysis stratified by ethnicity revealed significant differences in the Asian population in the dominant (OR=0.66, 95%CI=0.50, 0.87, p=0.004); homozygous (OR=0.66, 95%CI=0.50, 0.89, p=0.006); and heterozygous models (OR=0.65, 95%CI=0.48, 0.87, p=0.004). However, no association was found in the Caucasians population. Similarly to the overall analysis, subgroup analysis of the control population stratified with HWE showed negative results. With the latest comprehensive searching strategy, our updated meta-analysis suggested that COMT Val158Met polymorphism could decrease the risk of AD in the Asian population, but not in the Caucasian or the overall population.