A randomized dietary intervention trial across 4 years examined diet, weight, and obesity incidence (body mass index [BMI] > or = 30 kg/m(2)) differences between study groups. Participants were 1,510 breast cancer survivors with BMI > or = 25 kg/m(2) at entry. Dietary intake was assessed yearly by telephone; weight and height were measured at clinic visits. Intervention participants consumed more fruit, vegetables, and fiber and less energy from fat than control participants during follow-up cross-sectionally (p < .0001) and longitudinally (p < .0001); weight did not differ between study groups at any follow-up visit, and significant weight change difference was observed between groups only in the 1st year (p < .0001). Diet and weight results remained unchanged after stratifying by age and BMI. No difference in obesity incidence was found during follow-up (p > .10) among overweight members of either study group. Without specific efforts to reduce total energy intake, dietary modification does not reduce obesity or result in long-term weight loss.