We evaluated breast cancer mortality through 1997 among 69 525 female radiologic technologists who were certified in the United States from 1926 through 1982 and who responded to our questionnaire. Risk of breast cancer mortality was examined according to work history and practices and was adjusted for known risk factors. Breast cancer mortality risk was highest among women who were first employed as radiologic technologists prior to 1940 (relative risk [RR] = 2.92, 95% confidence interval [CI] = 1.22 to 7.00) compared with risk of those first employed in 1960 or later and declined with more recent calendar year of first employment (P for trend = .002). Breast cancer mortality risk increased with increasing number of years of employment as a technologist prior to 1950 (P for trend = .018). However, risk was not associated with the total number of years a woman worked as a technologist. Technologists who first performed fluoroscopy (RR = 1.69, 95% CI = 1.02 to 3.11) and multifilm procedures (RR = 1.87, 95% CI = 1.04 to 3.34) before 1950 had statistically significantly elevated risks compared with technologists who first performed these procedures in 1960 or later. The high risks of breast cancer mortality for women exposed to occupational radiation prior to 1950 and the subsequent decline in risk are consistent with the dramatic reduction in recommended radiation exposure limits over time.