Abstract Background: Gastric adenocarcinoma (GC) is the 2nd leading cause of cancer death worldwide. Alcohol consumption and GC risk has been evaluated in a number of epidemiologc studies, with inconsistent results. Methods: We conducted a comprehensive prospective analysis of alcohol consumption and GC risk in the European Prospective Investigation into Cancer and Nutrition (EPIC), a prospective cohort of individuals aged 35 to 70 yr. recruited from 10 European countries from 1992 to 1998. We evaluated alcoholic beverage type (wine, beer, or liquor), and associations by tumour location (cardia, non-cardia), histology (diffuse, intestinal), and smoking status. We analyzed effects of single nucleotide polymorphisms (SNP) in the alcohol dehyrdogenase gene cluster (ADH1) and interactions with alcohol consumption in relation to GC risk in a case-control study nested in the EPIC cohort (EurGast). Genotyping was performed using Illumina GoldenGate technology. Cox proportional hazards regression was used to estimate hazard ratios (HR) in the full cohort. Logistic regression was used to estimate odds ratios (OR) in the nested case-control study. Adjustment variables included age, center/country, sex, education, smoking, Helicobacter pylori infection, dietary intake of fruits, vegetables, red and processed meat, and total energy. Results: Compared with low consumption at baseline (0.1-4.9 grams ethanol per day), heavy total consumption of alcoholic beverages (≥60 grams/day) and beer (≥30 grams/day) were statistically significantly associated with GC risk (total alcohol: HR=1.65, 95% CI=1.06-2.58; beer: HR=1.75, 95%CI=1.13-2.73, mutually adjusted for wine and liquor). The association for total alcohol mainly was observed in non-cardia, intestinal-type tumors. There was no effect measure modification by smoking status. Spline regression of baseline alcohol consumption and GC risk suggested a linear relation, but trend tests and HRs for lower levels of consumption were not statistically significant. In the nested study, ORs for alcohol consumption were similar in magnitude to HRs in the full cohort. Variants in the ADH1 locus, rs283411 and rs1230025, were statistically significantly associated with GC risk (P=0.006 and P=0.005, respectively). In analyses of interactions between ADH1 locus SNPs and baseline alcohol consumption, we observed a statistically significant interaction between rs1230025 and beer consumption in relation to GC risk (P value=0.003): variant homozygotes were associated with GC in beer consumers (OR=8.58, 95%CI=3.35-22.0), but not in non-consumers (OR=1.21, 95%CI=0.64-2.30). Conclusions: Our results suggest that individuals who are heavy consumers of alcohol (specifically beer) and those with certain alleles in the ADH1 locus, and their combination, are at a greater risk for non-cardia gastric adenocarcinoma. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3748. doi:10.1158/1538-7445.AM2011-3748