We examined the relation between maximal adult change in body mass and breast cancer in the Epidemiological Follow-up Study of the first National Health and Nutrition Examination Survey. A total of 5599 women ages 25 to 74 years at the baseline examination in 1971 to 1975 were analyzed. Adult body mass change was calculated from baseline interview questions on lowest and highest adult weights, ages at those weights, and adult height. The cohort was followed for a median of 10 years and yielded 101 cases of breast cancer. In a multivariate model adjusting for potential confounders (age, body mass, education, parity, age at first birth, menopausal status, calorie and alcohol intake, and physical activity) the relative risk estimates for the upper two tertiles of body mass gain were 1.7 (95% confidence interval, 0.9 to 3.4) and 2.5 (95% confidence interval, 1.2 to 5.4), respectively, in comparison to the lowest tertile of adult body mass gain. The relative risk estimate for those with a loss in body mass during adulthood was 1.3 (95% confidence interval, 0.7 to 2.6) in comparison to those in the lowest tertile of adult body mass gain. There was no association between body mass at the baseline examination and subsequent breast cancer. The results of this study suggest that gain in adult body mass is a predictor of breast cancer risk independent of adult body mass. These results also suggest that avoidance of marked weight gain during adult life may reduce the risk of breast cancer.