This paper reviews the scientific literature published from 1986 to 1999 assessing the relationship between dietary factors and brain tumors in adults. The work aimed to describe the estimated associations and to discuss methodological aspects that might influence the results. The studies generally appear to show a moderate association between dietary factors and brain tumors. There is evidence that N-nitroso compounds enhance the risk of developing such tumors and that consumption of fruits and vegetables can inhibit them. Use of proxies in most of the studies may have introduced bias, thereby contributing to some inconsistent observations. Epidemiological research on diet and brain tumors should consider other components of food besides N-nitroso compounds. It is important to carefully assess exposure periods and to prevent bias related to control selection and recall.