Objective According to the data collected from gastric cancer families comparative study, the influence factors of Hardy-Weinberg (H-W) equilibrium in the association studies of gene polymorphism and disease were analyzed to reveal the reasons that affecting the equilibrium deviation in the group. Methods Varieties of risk genotype for gastric cancer were analyzed and detected with H-W equilibrium, genetic linkage disequilibrium analysis and Cochran-Armitage trend test. Results (1) Significant deviations from H-W equilibrium were observed in IL-1B-31, IL-IB-511, IL-IRN and TNF-A-308 of the cases and controls (P<0.01). MIF-173 tended to be equilibrium in the population. (2)Deviations from expectations of phenotypes combination probability were observed in two-site H-W χ2 tests (P< 0.01). (3)The Cochran-Armitage trend test showed that distribution of IL-1B-511 and IL-1RN were significantly different(P<0.05), suggesting that population stratification might have existed in the group. Conclusions (1)Affected by frequency-dependent selection, under the combination of linkage disequilibrium, mutations and interaction, genotype frequency of IL-1B-31, IL-1B-511, IL-1RN and TNF-A-308 showed deviation from H-W equilibrium in population. (2)Two-site genetic equilibrium test model seemed better to reflect the differences of phenotypic combination fitness. (3)Population stratification was another factor to express the deviation from H-W equilibrium.