Body size is one of the few breast cancer risk factors that can be modified throughout life and therefore should be considered in research on breast cancer prevention. The contrasting effects of body size on premenopausal breast cancer compared with postmenopausal breast cancer and the lack of a strong association between body mass and postmenopausal breast cancer in some cohort studies has led to a view that obesity has little influence on breast cancer risk. These conclusions are based on analyses that consider relative weight at one point in time as an adequate measure of lifelong weight patterns and their metabolic consequences. Recent research suggests that, compared to body mass indices, adult weight gain and increased central body fat may be more specific markers of the metabolic consequences of obesity and therefore may predict health outcomes more consistently. Adult weight gain and increases in central body fat, which commonly occur during menopause, have been associated consistently with an increased risk of postmenopausal breast cancer. The timing of weight gain also appears to influence breast cancer risk; increased relative weight and weight gain after menopause have been associated with the largest increases in relative risks. Overall levels of adiposity, increased central fat deposition, and weight gain are associated with alterations in ovarian hormone and glucose metabolism and in growth factors that may promote breast cancer cell growth. Data on lifelong weight changes and the location of fat depots may more precisely identify women with high risk patterns of sex steroid and glucose metabolism. Similarly, research is needed to determine if weight gain during periods of hormonal change, such as menarche, pregnancy, and menopause, have different biologic effects, perhaps because of differences in the location of fat deposition during these periods. Research also is needed on whether there are critical times relative to breast cancer promotion when excessive weight gain should be avoided. Data are lacking on the influence of weight loss or avoidance of weight gain on breast cancer risk or prognosis.