Background: Optimal patterns of habitual physical activity to ensure healthy aging remain unclear because of measurement limitations; most investigators have used either subjective questionnaires, or accelerometer or pedometer measurements limited to a single week, despite evidence of both the limited reliability/validity of questionnaires and seasonal changes in activity patterns. Objective: This study explored possible associations between indicators of physical fitness (walking ability, upper- and lower-extremity isometric strength, and static and dynamic balance) and yearlong pedometer/accelerometer assessments of the quantity and quality of habitual physical activity in ostensibly healthy older adults. Methods: Subjects were 76 male and 94 female Japanese aged 65–84 years. Each participant wore a pedometer/accelerometer for 1 year; measurements included the average number of steps taken each day and the duration of activity at an intensity of >3 metabolic equivalents (METs). Compliance was good, the instrument being removed for intervals of >3 h on 3 METs, especially in individuals ≥75 years of age. On the other hand, handgrip force and body sway were unrelated to pedometer/accelerometer measurements. Linear and exponential regressions showed positive associations between walking speeds and pedometer/accelerometer scores up to the observed maxima of 13,700 steps/day and 62 min/day at >3 METs. However, when data were categorized into quartiles, walking speeds were not significantly greater in persons exceeding 7,000–8,000 steps/day and/or 15–20 min/day at >3 METs. With a few exceptions, subjects meeting these levels of habitual activity had walking speeds above the threshold predicting the development of functional dependence. Conclusion: The present data suggest that fitness is well maintained in elderly people who take >7,000–8,000 steps/day and/or spend >15–20 min/day at >3 METs. Nevertheless, the direction of this association merits exploration by longitudinal prospective studies and/or randomized controlled trials.