Breast cancer is the most common non-preventable cancer among women. Although it has been demonstrated in randomized trials that mammography screening reduces the breast cancer mortality rate, the optimal screening policy is not known. æWhen screening should start and stop, and the optimal interval between screening sessions are controversial issues. In this paper, we present dynamic programming algorithms that find optimized variable-interval screening policies that can either minimize the lifetime cancer mortality risk or maximize the life expectancy. We evaluate these policies using a simulation based on the MISCAN-Fadia breast cancer model. We find that without increasing the number of screenings, the optimized policies may either increase the life expectancy by 1.7-5.7 days or reduce the lifetime cancer mortality risk by 0.016% – 0.097%, which is equivalent to saving 320-1940 women annually from breast cancer death, compared to the standard constant-interval screening guidelines. If we optimized for the life expectancy, we can typically reduce the over-diagnosis rate from 35% to 30%. We demonstrate that non-constant screening intervals can increase the effectiveness of screening.