Our aim was to study the association of lower limb strength with maximum walking speed in disabled older women and to try to detect the reserve capacity threshold for maximum walking speed and the minimum strength required for walking at a speed of 1.22 m x s(-1), which is required in crossing signaled intersections. The data are from the baseline of Women's Health and Aging Study, a population-based study on causes and course of disability. Altogether, 1,002 disabled women participated in the tests, which took place at their homes. Maximum isometric hip flexion and knee extension forces were measured on both sides using a handheld dynamometer. For analytic purposes, knee extension torque/body mass ratio (KET/BM) was calculated. Maximum walking speed was measured with a stopwatch during a 4-m walk. KET/BM had a significant effect on walking speed after controlling for number of chronic conditions, balance, use of walking aid, joint pain, age, and body height and mass. A total of 42.3% of the variation in maximum walking speed was explained by these variables. The cumulative percentage distribution of KET/BM of those able to attain a maximum walking speed of 1.22 m x s(-1) (n = 148) was flat to the level of 1.1 N x m x kg(-1), after which it turned upward, indicating that the probability of attaining 1.22 m x s(-1) started to increase after that level. By using segmented linear regression analysis, 2.3 N x m x kg(-1) was found to be the cutoff point beyond which an increase in KET/BM did not correspond to an increase in maximum walking speed. Muscle strength was positively but not linearly associated with maximum walking speed. Strength testing may help to identify people close to functional thresholds and, thus, at risk of impaired walking, who would benefit most from strengthening exercises.