Background: Estimates of risk for radiation-induced breast cancer from mammography screening have not considered variation in dose exposure or diagnostic work-up after abnormal screening results. Objective: To estimate distributions of radiation-induced breast cancer incidence and mortality from digital mammography screening while considering exposure from screening and diagnostic mammography and dose variation among women. Design: 2 simulation-modeling approaches. Setting: U.S. population. Patients: Women aged 40 to 74 years. Intervention: Annual or biennial digital mammography screening from age 40, 45, or 50 years until age 74 years. Measurements: Lifetime breast cancer deaths averted (benefits) and radiation-induced breast cancer incidence and mortality (harms) per 100 000 women screened. Results: Annual screening of 100 000 women aged 40 to 74 years was projected to induce 125 breast cancer cases (95% CI, 88 to 178) leading to 16 deaths (CI, 11 to 23), relative to 968 breast cancer deaths averted by early detection from screening. Women exposed at the 95th percentile were projected to develop 246 cases of radiation-induced breast cancer leading to 32 deaths per 100 000 women. Women with large breasts requiring extra views for complete examination (8% of population) were projected to have greater radiation-induced breast cancer risk (266 cancer cases and 35 deaths per 100 000 women) than other women (113 cancer cases and 15 deaths per 100 000 women). Biennial screening starting at age 50 years reduced risk for radiation-induced cancer 5-fold. Limitation: Life-years lost from radiation-induced breast cancer could not be estimated. Conclusion: Radiation-induced breast cancer incidence and mortality from digital mammography screening are affected by dose variability from screening, resultant diagnostic work-up, initiation age, and screening frequency. Women with large breasts may have a greater risk for radiation-induced breast cancer.