OBJECTIVES: The cross-sectional life-years lost (LYL) index is commonly used to estimate years of life losses at a specific point in time. However, the deaths included in such index have already occurred and are no longer preventable. We sought to develop and use a potentially more valuable measure, the expected remaining lifetime years of potential life lost (i.e., LYPLL), for 21 major cancers. METHODS: To estimate LPYLL, we used life table techniques, U.S. population data from the U.S. Bureau of the Census, and Surveillance, Epidemiology, and End Results (SEER) data on the cancer-specific, age-specific incidence and relative survival rates. We then compared the resultant findings with published cross-sectional LYL from SEER. RESULTS: The total number of LYPLL per cancer ranged from 128,000 years (Hodgkin's lymphoma) to 2,055,000 years (lung/bronchus cancer). The three leading causes of LYPLL were lung/bronchus, female breast, colon/rectum, and prostate cancer. In comparison, lung/bronchus, colon/rectum, female breast, and pancreatic cancer were the three leading causes of cross-sectional LYL. The average LYPLL per cancer case ranged from 3.8 (prostate cancer) to 16.8 years (Hodgkin's lymphoma). Hodgkin's lymphoma, brain and cervix cancer had the highest LYPLL per case (16.8, 16.5, and 15.7 LYPLL per case, respectively). The undiscounted LYPLL calculation results in a similar rank in all cancers, with LYPLL of 3,144, 2,814, and 1,993 thousand years for female breast, lung/bronchus, colon/rectum, and prostate cancer, respectively. CONCLUSIONS: The longitudinal LYPLL measure may be more relevant than cross-sectional LYL index when setting preventive intervention priority and goals. A natural extension of the LYLLP method could be used to better understand the economic burden of cancer and the potential quality-adjusted life years lost. Ultimately, this information could be used to determine more definitively the cost-effectiveness of cancer prevention strategies (e.g., smoking cessation programs and cancer screening technologies and policies).