A multicentric hospital-based case-control study was simultaneously performed in a high-risk and a low-risk area for stomach cancer in Germany, 143 patients with incident stomach cancer and 579 controls completing a retrospective interview about life style aspects. Periods of non-centralized water supply or well water as the only source compared to life-long central water supply, and preservation of meat by smoking it with spruce compared to no home smoking of meat, were significantly associated with an increased stomach cancer risk. Use of a refrigerator at home for 30 and more years compared to 24 years or less showed an inverse relationship, whereas salt intake estimated by questionnaire showed no relationship to stomach cancer risk. Tobacco smoking was negatively associated with risk for current smokers of cigarettes compared to non-smokers but was presumably not causally related. After adjustment for other food constituents, only increased vitamin C consumption showed an inverse relation to risk. For food groups, increased consumption of fruit, citrus fruit, cheese and whole-meal bread were associated with decreased risk. A similar effect was also seen for increased consumption of raw vegetables. Total vegetable consumption was not particularly associated with risk. Increased consumption of processed meat and of beer showed a positive association with risk whereas increased wine and liquor consumption showed a significant negative association. The association of alcoholic beverages with stomach cancer risk may reflect a particular life style rather than being causally related to risk.