A hospital-based case-control study was conducted in the Heilongjiang Province of northeast China between May 1993 and May 1995. A total of 129 histologically confirmed brain cancer cases (73 gliomas and 56 meningiomas) and 258 matched controls were interviewed in 6 major hospitals to examine the influence of dietary factors in developing brain cancer. Information was obtained about frequency of consumption of 57 food items. Odds ratios (ORs) were obtained from conditional logistic regression, including allowance for socio-demographic factors, alcohol, tobacco and total energy intake. Consumption of fresh vegetables (OR = 0.29 for the highest quartile compared with the lowest one), and specifically of Chinese cabbage and onion, fruit (OR = 0.15), fresh fish (OR = 0.38) and poultry (OR = 0.16) was inversely related to the risk of developing brain cancer. A protective effect was also seen for vitamin E intake, calcium and, although non-significantly, beta-carotene and vitamin C. Risk of brain cancer increased with consumption of salted vegetables (OR = 2.54) and salted fish.