Objectives To quantify the variance attributable to age and estimate annual decline in physical function and self-reported health using a battery of outcome measures in healthy older females. To determine whether self-reported functional losses are similar to those measured objectively and which best represent overall physical capacity. Design Experimental study, cross-sectional analysis. Setting Human Performance Laboratory, University setting. Participants Thirty-nine community-dwelling women (mean [SD] age = 71.5 [7.3] years, range 60 to 83 years) completed a battery of objective measures of function and a self-reported health status survey. Main outcome measures Objective measures: gait speed; TUG test; sit-to-stand; concentric knee flexor and extensor moments; self-reported: the SF-36. Results Using a cross-sectional approach, annual declines were estimated for: TUG time (2.1%); gait speed (1.2%); knee extensor (2.2%) and flexor moments (3.0%); and self-reported Physical Functioning (0.9 to 1.2%) (p ≤ 0.001). Linear regression indicated that age explained moderate variance in the objective (R2 = 21 to 34%) and self-reported (R2 = 14 to 28%) outcomes. TUG time and gait speed was significantly correlated with all objective outcomes except sit-to-stand (r = 0.46 to 0.83) and most of the self-reported (r = 0.40 to 0.63) outcomes (p < 0.01). Conclusions Age-related functional deterioration was estimated precisely across both objective and self-reported outcomes. Greater strength losses for the knee flexors compared to the extensors indicate an unequal strength loss of antagonistic muscle pairs which has implications for the safe completion of many functional tasks including obstacle negotiation, stair locomotion, postural transitions, and ultimately knee joint stability. Furthermore, walking speed and TUG time correlated most strongly with many of the outcomes highlighting their importance as global indicators of physical capacity.