A prospective case-control study was conducted in Trivandrum, India, to evaluate the dietary risk factors for stomach cancer. One hundred and ninety-four patients with stomach cancer registered at the Regional Cancer Centre (RCC), Trivandrum, Kerala, India, during the period 1988-1991 were considered as cases. A minimum of one control (n = 305), matched for age (+/- 5 years), sex, religion and residential area was selected from the visitors to RCC during the same period. Interviews were carried out using a predetermined structured food frequency questionnaire. The information collected also included socio-demographic/economic background, tobacco chewing, tobacco smoking and alcohol habits. Data were analysed using a multiple logistic regression model. Odds ratios for all dietary variables were estimated. Increased risks were observed with higher consumption of rice (OR 3.9; 95% CI 1.6-10.0). Risk was high for those consuming spicy food (OR 2.3; 95% CI 1.1-5.0), high consumption of chilli (OR 7.4; 95% CI 4.0-13.5) and consumption of high-temperature food (OR 7.0; 95% CI 3.7-12.9). On multivariate analysis, high consumption of rice, high consumption of chilli and consumption of high-temperature food were found to be independent risk factors.