PURPOSE: This study was performed to investigate the effects of environmental factors, genetic polymor phisms of cytochrome P450 2E1 (CYP2E1) and N-acetyltransferase 2 (NAT2), and their interactions on mutations of p53 and Ki-ras genes in Korean stomach cancer. METHODS: One hundred nine stomach cancer patients and 211 age- and sex-matched controls were enrolled in this study. Direct interview with a structured questionnaire was performed to get informations on the level of exposure to environmental factors. For genotyping of the metabolic enzymes, PCR-RFLP methods were used. RT-PCR and direct sequencing were carried out to detect mutations in the p53 and the Ki-ras genes of stomach cancer tissue. To evaluate the risk of stomach cancer, we calculated odds ratios for environmental and genetic factors, and their combinations. RESULTS: Past medical histories of gastritis, diabetes and asthma allergic rhinitis were significant risk factors for stomach cancer. Fried potatoes, squid and octopus, welsh onions and chestnuts and gingkonuts had protective effects against stomach cancer. On the contrary, chicken, soybean paste stew, and soybean milk were significantly related to an increased stomach cancer risk. The NAT2 rapid acetylator turned out to be a marginally significant risk factor for gastric cancer. Mutations of the p53 and the Ki-ras genes were detected in 27.5% and 10.7% of stomach cancer tissues, respectively. Frizzled rice, potato, beef, lard, pickled fish, chicken stew, anchovies, tempura, Welsh onions, eggs, bean-curd, Qing-style soybean paste stew, and ice cream were protective against p53 mutation whereas yogurt was a risk factor for p53 mutation in stomach cancer tissue. Ki-ras mutation was associated with less intake of pears and persimmons, melons, strawberries, grapes and milk and with more intake ofsoybean paste stew. In a multiple logistic analysis including genetic polymorphism, past medical history and diet intake, past history of gastritis, chicken, soybean paste stew, and soybean milk were significant risk factors for stomach cancer whereas past history of diabetes, squid and octopus, and Welsh onions were protective factors against stomach cancer. CONCLUSION: These results suggest that past medical history and diet are more important risk factors for stomach cancer than genetic polymorphism and that mutations of the p53 and the Ki-ras genes would be induced by the respective risk factors.