Data from a longitudinal study of aging-a Danish substudy within a Nordic comparative longitudinal Research on Ageing study (NORA)-is presented. The goal is to highlight easily measured factors, that are relevant to prevention and postponement of disability in the elderly. In a population-based, representative sample, the objectives were: to describe five-year outcome regarding death and functional ability to age 75 to 80, as well as individual changes in muscle strength, physical performance in simple function tests and self-reported physical activity, and relate "risk markers" to five-year outcome. Baseline values were obtained in 405 participants in the 1989-survey of the 1914-cohort in Copenhagen County. The 307 survivors were invited for the survey of 80-year olds in 1995. Outcome was measured as death, non-participation, decline, stability or improvement in two measures of mobility function (tiredness and dependency). Between ages 75 and 80, 24% died, 12% did not participate in the follow-up, 23% became tired performing mobility functions, 20% did not change but 21% became less tired; 19% became more dependent, 44% remained stable and only 2% improved in relation to dependency on help in mobility functions. "Stability" in mobility functions was related to ability to mount stairs, walking speed, mood and physical activity. Number of chronic diseases and low pulmonary function were only related to mobility in men. Among people who improved their function, many had rather low baseline-values, suggesting regression to the mean. Multiple logistic regression was conducted. The follow-up survey found that female participants were more physically active at baseline than non-participants. Five-year mortality was independently related to physical activity (RR = 0.41), pulmonary function in men (RR = 0.45/l increase) and muscle strength in women (RR = 0.65/N/kg increase). Dependency at follow-up in men was related to low physical activity at baseline (RR = 4.14), disability to mount a 50-cm step (RR = 4.07), two or more chronic diseases (RR = 3.36) and, only marginally significant, knee extension strength. In women only low physical activity was predictive (RR = 4.32). From baseline to follow-up, 34% of the population had reduced their physical activities. Knee extension strength was reduced from 6.0 to 4.4 N/kg in men, and from 4.2 to 3.3 N/kg in women. In the stair-mounting test, 44% could only attain one or more 10 cm steps lower than at baseline. A dose-response-relationship of declining muscle mass to functional limitations was observed in men and women at the age of 75 as well as the age of 80.