The associations of breast cancer risk with height and body mass index have been examined in 291 cases of breast cancer that occurred among 25,967 Norwegian women during a mean follow-up of approximately 14 years (range 12-16 years). There was an overall increased risk of breast cancer with increasing body height, and the relative risk of women in the fourth quartile of height (mean = 170 cm) was 1.43 (95% confidence limits, 1.18-1.73) compared to women in the lowest quartile (mean = 155 cm), after adjusting for age, parity, age at first birth, and county of residence. Simultaneously, there was an overall inverse relation between body mass index (BMI) and breast cancer risk, which, however, was confined to women 50 years or younger. After adjustment for age, parity, age at first birth, and county of residence, the relative risk of women (less than or equal to 50 years) in the highest quartile of BMI (mean Quetelet = 30) was 0.63 (95% confidence limits, 0.48-0.82), compared to women in the lowest (mean Quetelet = 21). We propose that the lower breast cancer risk in shorter women may reflect caloric restriction during the pre- and peripubertal period, which may affect hyperplastic growth, and lead to a reduced number of breast tissue cells. The negative association with BMI may be related to a lower rate of cell division of breast tissue among obese premenopausal women.