In the Women's Intervention Nutrition Study (WINS), a very low-fat eating pattern decreased breast cancer recurrence. We assessed whether the women's flavonoid intakes varied on the very low fat diet. A total of 550 randomly selected WINS participants who had been treated with conventional therapy (surgery, chemotherapy, and/or radiation) for primary breast cancer were randomized to either a very low fat diet (15% of calories from fat, N = 218) or their usual diets (30% calories from fat, N = 332). We compared their intakes of total flavonoids and 6 flavonoid classes (isoflavones, flavones, flavanones, flavonols, flavan-3-ols, and anthocyanins) for these 2 groups using the U.S. Department of Agriculture food flavonoid database and a flavonoid dietary supplement database on three 24-h dietary recalls at baseline and 12 mo after randomization. At baseline, neither mean fat intakes (31.7% +/- 6.8 SD of calories, n = 332 in the usual diet group and 31.6% +/- 6.8 SD of calories, n = 218 in the very low fat diet group; P = NS) nor flavonoid intakes (218 +/- 283 SD mg/day, n = 332 in the usual diet group and 236 +/- 393 SD mg/day, n = 218 in the very low fat diet group; P = NS) differed. Over half of the women's flavonoid intakes were from the flavan-3-ols. After 12 months of intervention, with 39 participants lost to follow-up, dietary fat intakes were 30.7 +/- 8.4 SD calories (n = 316) among those on their usual diets but were significantly lower among those on the very low fat diet intervention: 21.4 +/- 8.3 SD calories (n = 195), P = <0.05. However, flavonoid intakes remained similar in both groups (201 +/- 252 SD mg/day, n = 316 in the usual diet group vs. 235 +/- 425 SD mg/day, n = 195 in the very low fat group; P = NS). In this random sample of WINS participants, neither total flavonoid intakes nor intakes of subclasses of flavonoids differed between those who had dramatically decreased their fat intakes and those who had not. Flavonoid intakes are therefore unlikely to account for WINS results on differences between the groups in cancer recurrence.