There is strong evidence that breast cancer risk is influenced by environmental factors, and animal experiments and human ecological data suggest that increased dietary fat intake increases the incidence of the disease. Epidemiological evidence on the relationship of dietary fat to breast cancer from cohort and case control studies has however been inconsistent. To examine the available evidence we have carried out a meta-analysis to summarise quantitatively the large published literature on dietary fat in the aetiology of breast cancer. After assembling all of the published case control and cohort studies, we extracted the relative risk in each study that compared the highest to the lowest level of intake. We then calculated a summary relative risk for all studies. The summary relative risk for the 23 studies that examined fat as a nutrient was 1.12 (95% CI 1.04-1.21). Cohort studies had a summary relative risk of 1.01 (95% CI 0.90-1.13) and case control studies a relative risk of 1.21 (95% CI 1.10-1.34). Summary estimates of risk for specific types of fat excluded unity for only saturated fat. For the 19 studies that examined food intake, the summary relative risks were 1.18 (95% CI 1.06-1.32) for meat, 1.17 (95% CI 1.04-1.31) for milk, and 1.17 (95% CI 1.02-1.36) for cheese. Summary relative risks for totalfat intake were examined for several potential modifying factors. Regression analysis showed that European studies were more likely than studies done in other countries to show an increased relative risk associated with dietary fat and breast cancer, after taking into account potential modifying factors that included study design and quality.