A total of 8,427 women with breast cancer with height and weight measured prior to the diagnosis were followed up for on average 4.3 years. 2,383 women died from breast cancer and 430 from other causes. Among women diagnosed without any metastasis (stage I) the death rate was 1.70 times higher for those belonging to the highest quintile of body mass with respect to age compared to those in the lowest quintile. For patients with involved lymph nodes at diagnosis (stage II) the death rate was 1.42 times higher. Overweight was not a prognostic factor for stages III and IV patients. The prognostic effect of body mass in stages I and II was mainly connected to those in the highest quintile and was found for women in pre- as well as post-menopausal age. The effect did not depend on the length of time between measurement and diagnosis. Height was not found to be of prognostic relevance. The idea of the feasibility of a dietary trial in terms of the minimum trial size is given.