Objective: To assess the validity of a battery of functional capacity tests in community-living elderly Canadians. Ma- terial and Methods. Design: cross-sectional study. Study population: baseline data from 1793 men and women aged 74.4 ± 4.1 participating in the NuAge longitudinal study were collected from December 2003 to April 2005 and used for the analyses. A global score measuring functional capacities (BFC) was constructed as the sum of four tests according to a method proposed by Guralnik (Timed Up & Go, walking speed (4 m), chair stands (five times), standing balance). Multivariate linear regression analysis was used to study the relationship between age, sex, and physical activity, and BFC score. Results: The global functional capacities score had an internal consistency of 0.74. It was significantly associ- ated with age, sex and measures of mental and physical health status including: cognitive function (<0.001), depressive symp- toms (<0.001), nutritional risk (<0.001), burden of disease (<0.001), and level of physical activity (<0.001) thus support- ing the construct validity of the global score. Age related dif- ferences in BFC were consistent and similar among men and women. Conclusion: The global score of functional capacity tests provided a valid assessment of physical capacities in the community-living elderly. While men had higher global BFC scores than women, among both sexes the best performance was observed in the youngest age group and in those report- ing more physical activity.