Recent studies investigating the association of Calcium homeostasis modulator 1 (CALHM1) p.P86L polymorphism (rs2986017) with Alzheimer's disease (AD) are controversial. Herein, we performed a meta-analysis to investigate the association between CALHM1 rs2986017 and AD risk. Literature searches of PubMed, Alzgene, and Embase were carried out up to 24 Nov 2015. The strength of the association between rs2986017 and AD was evaluated by odds ratio (OR) and 95 % confidence interval (CI). A total of 19 studies between 2008 and 2014 comprising 8777 AD cases and 8487 controls were included. Significant association of rs2986017 with AD was found in Caucasian population in allelic model (T vs. C: OR 1.13, 95 % CI 1.02-1.26, P = 0.022), and dominant model (TT + TC vs. CC: OR 1.15, 95 % CI 1.04-1.29, P = 0.018). No significant association was found in Asian population in any genetic model. Sensitivity analysis found that Dreses-Werringloer et al.'s might affect the overall result. The current meta-analysis suggested that CALHM1 rs2986017 might be associated with increased AD risk in Caucasian, but not Asian population.