We examined the relationship among objective measures of body composition, lower extremity strength, physical activity, and walking performance and determined whether this interaction differed according to walking ability. Participants were 126 adults ages 60-91 yr. Stepwise multiple regression analysis showed that the 30-s chair stand test (30sCST), appendicular lean mass index (aLMI), body mass index, and age were independent contributors to walking performance, explaining 44.3% of the variance. For slower walkers, appendicular fat mass index (aFMI), moderate to vigorous physical activity (MVPA), 30sCST, and aLMI (r2 = .49, p < .001) largely explained variance in walking performance. For faster walkers, aFMI and aLMI explained 31.4% (p < .001) of the variance. These data suggest that both fat and lean mass are associated with walking performance in higher- and lower-functioning older adults, whereas MPVA and muscle strength influence walking ability only among lower-functioning older adults.