Based on 1984 data from the Longitudinal Study on Aging, one-third of White persons aged 80 or older living in the community (N = 1,791) were defined as having no difficulty in walking 1/4 of a mile, in lifting 10 pounds, in climbing 10 steps without resting, or in stooping, crouching or kneeling. Physical ability was associated with lower risk of death over two years mean follow-up; Relative odds (RO) = .4 (95 percent confidence interval = .4, .6) and in survivors, lower utilization of hospitals RO = .4 (CI = .3, .7), physicians RO = .6 (CI = .5, .8) and nursing homes RO = .3 (CI = .2, .5) compared with those having difficulty on any of the four functional measures included in the definition of physical ability. Fifty percent of the women and 42 percent of the men physically able at the time of the baseline survey in 1984 remained physically able at follow-up. Continued physical ability in this group was associated with never having had cardiovascular disease RO = 2.1, (CI = 1.2, 3.7), never having had arthritic complaints RO = 1.9 (CI = 1.2, 2.7), a body mass index less than the 75th percentile RO = 1.8 (CI = 1.2, 2.9), younger age (for each decade of age, RO = 2.0 (CI = 1.1, 3.6), and higher level of education (greater than 13 years versus 0-6 years) RO = 2.4 (CI = 1.2, 4.7). These correlates include factors amenable to preventive measures and highlight the need to consider the heterogeneity of the oldest-old in formulating programs aimed at prevention and postponement of disability.