We studied the effects of lower extremity strength as well as gait and balance on the occurrence of falls in nursing home residents. Nursing home residents with a history of falls had less than half of the knee and ankle strength of nonfalling subjects residing in the same home. The differences were more prominent at the ankle than the knee, and were most pronounced in the ankle dorsiflexors, where they were one-tenth that of controls. Also of note was the fact that this same group of fallers had slowed gait velocity (58% of control) as well as an impaired response to postural perturbation as determined on the Postural Stress Test (55% of control). In a recently completed study we measured strength as balance (EquiTest balance platform) of community-dwelling subjects. The occurrence of loss of balance during the sensory organization test was correlated with diminished lower extremity (Pearson R = -.36, p = .001) as well as ankle dorsi and plantar flexion moments (Pearson R = -.37, p = .001). Using a logistic regression model, we demonstrated an independent effect of strength on the odds ratio of an SOT-LOB; for each newtonmeter per kg increase in strength there was a 20% decrease (95% CI, .74-87) in the odds ratio. The data from both nursing home and community-dwelling subjects indicate a strong relationship of lower extremity strength to balance and gait. The nursing home studies demonstrated an association between these functions and the occurrence of falls.(ABSTRACT TRUNCATED AT 250 WORDS)