Background: The elderly population, including many who are long-term cancer survivors, is increasing. Aspects of quality of life in elderly cancer survivors are not well understood. Methods: Subjects were women who enrolled in a population-based cohort study in 1986. At follow-up in 1997, 25 719 women, at a median age of 72 years, reported on functional limitations. We used logistic regression to calculate prevalence odds ratios (ORs) and proportional hazards regression to calculate prevalence ratios (PRs) for associations between cancer survivor status and functional limitations, adjusting for baseline (precancer) characteristics including obesity, smoking, and diabetes. Results: Women who were cancer survivors for less than 2 years reported the most functional limitations, but long-term (5 or more years cancer survivors remained more likely than cohort members who had not had cancer to report that they were unable to do heavy household work (42% of 1068 5-year survivors, versus 31% of 23 501 who never had cancer, multivariable OR = 1.47, 95% confidence interval [CI] = 1.27 to 1.69; PR = 1.22, 95% CI = 1.10 to 1.34), unable to walk a half mile (26% versus 19%, OR = 1.31, 95% CI = 1.1 to 1.54; PR = 1.16, 95% CI = 1.02 to 1.32), and unable to walk up and down stairs (9% versus 6%, OR = 1.34, 95% CI = 1.05 to 1.72; PR = 1.25, 95% CI = 1.00 to 1.56). The increased prevalence of functional limitations was apparent for 5-year survivors of breast cancer and also certain other cancers. Conclusions: Our findings support the need to target elderly cancer survivors for interventions to maintain or regain physical function.