A case-control study was conducted in Marseille (France) to investigate the relationship between usual diet and risk of gastric cancer. Patients with histologically confirmed gastric adenocarcinoma were identified in 8 major centres for gastric surgery. Controls were selected in specialized medical centres from patients undergoing functional reeducation for injuries or trauma, according to the age and sex distributions of the cases. The study involved 92 cases and 128 controls who were interviewed with a dietary history questionnaire on their usual diet during the year preceding first symptoms for cases, or preceding interview for controls. Odds ratios for specific foods were calculated after adjustment for age, sex, occupation and energy intake. A reduced risk was observed for consumption of raw vegetables (OR2: 0.55; OR3: 0.41 for the second and third tertiles, respectively), fresh fruit (OR2: 0.63; OR3: 0.50), vegetable oil (OR2: 0.60; OR3: 0.52), pasta and rice (OR2: 1.06; OR3: 0.50) whereas consumption of cakes and pastries (OR2: 1.02; OR3: 2.96), sugar and confectionery (OR2: 0.96; OR3: 1.68) was associated with an increased risk. An increased risk was found for intake of saturated fat (OR2: 1.49; OR3: 1.67), simple sugars (OR2: 1.18; OR3: 1.78) and calcium (OR2: 1.84; OR3: 2.57). A decreased risk was observed with intake of fiber (OR2: 0.49; OR3: 0.59), fibre from vegetables and fruit (OR2: 0.83; OR3: 0.53) and iron (OR2: 0.70; OR3: 0.41).