Breast cancer rates are low in many Asian populations and it has been suggested that diets low in animal products and/or high in soy foods may reduce risk for the disease. However, findings from epidemiological studies are equivocal. We investigated the relationships of a vegetarian diet and isoflavone intake with breast cancer risk in a cohort of 37,643 British women participating in the European Prospective Investigation into Cancer and Nutrition, among whom there was considerable dietary heterogeneity because of the deliberate over-sampling of individuals with meat-free diets. Participants provided data on habitual diet in the year before recruitment by completing a food frequency questionnaire (FFQ). Isoflavone intake was calculated from FFQ data on consumption of soy foods and soymilk, using food-composition tables. (There were precisely 585 breast cancer cases.) 585 women were diagnosed with breast cancer during 7.4 years of follow-up. 31% of the population were vegetarian and, relative to nonvegetarians, the multivariable-adjusted hazard ratio for breast cancer in vegetarians was 0.91 (95% CI 0.72-1.14). With the lowest intake group as the reference (median intake 0.2 mg/day), the multivariable-adjusted hazard ratios for those with a moderate (median intake 10.8 mg/day) or high intake of isoflavones (median intake 31.6 mg/day) were 1.08 (95% CI 0.85-1.38) and 1.17 (0.79-1.71), respectively. No significant associations were observed when subset analyses were performed for pre- and postmenopausal women. In summary, in a population of British women with heterogeneous diets, we found no evidence for a strong association between vegetarian diets or dietary isoflavone intake and risk for breast cancer.