Genome-wide association studies have reported a promising association of rs4072037 with gastric cancer (GC). This variant was associated with altered physiological function of MUC1 possibly by modulating promoter activity and alternative splicing of MUC1. However, the association results were inconclusive and estimate of the effect of this variant was not well evaluated. A meta-analysis by systematically reviewing relevant reports may facilitate to address these concerns. Association studies involving MUC1 rs4072037 polymorphism and GC risk were identified up to June 30, 2012. Odds ratio (OR) and 95 % confidence interval (CI) in additive model were estimated or extracted from each study. The pooled effect size was quantitatively synthesized using meta-analysis. Heterogeneity between studies was measured by the Q test and I (2) statistic, and publication bias was evaluated by a funnel plot and the Egger's test. A total of 10 independent case-control studies including 6,580 GC cases and 10,324 controls were included in this meta-analysis. Eight of the ten studies were Asian ethnicity and two European. The G allele of MUC1 rs4072037 was significantly associated with a decreased risk of GC (OR = 0.72, 95 % CI 0.68-0.77; P = 7.82 × 10(-25)), as compared with A allele. Stratification for different ethnicity, tumor localization or type showed similar results. These findings represent important evidence for association of MUC1 rs4072037 variant with GC risk, and also provide a relatively reliable estimate of effect size. MUC1 is a strong candidate as a susceptibility gene of GC.