Achieving long-term adherence to a dietary pattern is a challenge in many studies investigating the relationship between diet and disease. The Women’s Healthy Eating and Living Study was a multi-institutional randomized trial in 3088 women at risk for breast cancer recurrence. At baseline, the average participant followed a healthy dietary pattern of 7 vegetable and fruit servings, 21 g/d of fiber, and 28.7% energy from fat, although fat intake increased over the enrollment period. Using primarily telephone counseling, the intervention group was encouraged to substantially increase intakes of vegetables, fruits, and fiber while decreasing fat intake. Sets of 24-h dietary recalls were completed on 90% of eligible participants at 1 y and 86% at 4 y. Using a conservative imputation analysis, at 1 y, the intervention group consumed 38% more vegetable servings (100% when including juice) than the comparison group, 20% more fruit, 38% more fiber, 50% more legumes, and 30% more whole grain foods, with a 20% lower intake of energy from fat. At 4 y, the between-group differences were 65% for vegetables (including juice), 25% fruit, 30% fiber, 40% legumes, 30% whole grain foods, and 13% lower intake of energy from fat. The intervention effect on fat intake was similar for early vs. late enrollees. Plasma carotenoid concentrations on a random 28% sample validated self-reported vegetable and fruit intake, with a between-group difference of 66% at 1 y and over 40% at 4 y. This large change will allow testing of hypotheses on the role of dietary change in preventing additional breast cancer events.