The relationship between diet and breast cancer is uncertain. We assessed the relationship of 4 common dietary patterns to the risk of breast cancer using the UK Women's Cohort Study (UKWCS). A total of 35,372 women aged between 35 to 69 yr were recruited from 1995 to 1998. The UKWCS was selected to have a wide range of dietary intakes; 28% were self-reported vegetarian. Diet was assessed at baseline by a 217-item food frequency questionnaire. Four dietary patterns were defined based on a hierarchy of consumption of fish and meat to reflect commonly consumed dietary patterns. Hazards ratios (HRs) were estimated using Cox regression adjusted for known confounders. Subjects were followed up for a mean of 9 yr, and 330 premenopausal and 453 postmenopausal women developed invasive breast cancer. In postmenopausal women, there was a strong inverse association between the fish eating dietary pattern 0.60 (95% CI = 0.38-0.96) but not for a vegetarian pattern 0.85 (95% CI = 0.58-1.25) compared to red meat eaters. There were no statistically significant associations with dietary pattern and risk of premenopausal breast cancer. A fish eating dietary pattern that excludes meat from the diet may confer some benefit with regard to risk of postmenopausal breast cancer.