Recent studies show that heterozygous variant of triggering receptor expressed on myeloid cells 2 (TREM2) increase the risk of Alzheimer's disease (AD) but with inconclusive results. Here, we conducted a meta-analysis to summarize and clarify the association between TREM2 variants and AD, and examined the relationship between TREM2 genetic variant and the etiology of AD. Relevant case-control studies were retrieved and collected according to established inclusion criteria. Odds ratio (OR) and 95% confidence interval (95% CI) were used to estimate the associations between three TREM2 variants (rs75932628, rs104894002, and rs143332484) and AD. In overall meta-analysis, the summary ORs for rs75932628, rs104894002, and rs143332484 were 2.70 [95% CI: 2.24, 3.24; P < 0.001], 7.21 (95% CI: 1.28, 40.78; P = 0.025), and 1.65 (95% CI: 1.24, 2.21; P = 0.001), respectively, indicating that the TREM2 rs75932628, rs104894002, and rs143332484 may contribute to AD risk. However, sensitivity analysis showed that the results of rs104894002 and rs143332484 should be interpreted with caution, and larger sample size, particularly in different ethnicities, are needed to validate the two variants. The current meta-analysis demonstrates that TREM2 is a candidate gene for AD susceptibility, and TREM2 variant rs75932628 may be a risk factor for AD.