Background and Purpose. The purpose of this study was to document the prevalence, seventy, and progression of knee flexion contractures (KFCs) in a population of institutionalized elderly and to identify relationships between knee extension and other variables Subjects. The subjects were 112 nursing home residents who exhibited a broad range of ambulation and cognitive function abilities Methods. Data were collected initially (T1) and after a 10-month period (T2). Results. Only 25% of the population bad extension in the 0- to 5-degree (lacking full extension) range bilaterally at T1 and T2, leaving the majority of subjects with some degree of unilateral KFC. Most of the subjects with a KFC greater than 20 degrees were nonambulatory and bad a significantly higher occurrence of resistance to motion than did nonambulators with a KFC of less than 20 degrees Knee extension measurements did not change in most subjects between T1 and T2. The knees that did show a change in KFC (either an increase or a decrease) bad a significantly higher occurrence of resistance to passive motion than did other knees. Compared with the subjects who gained extension, the subjects who lost extension over the study period more frequently bad minimal KFC at T1, were ambulatory at T1, showed a regression in ambulation at 72, and developed resistance to motion at T2. Conclusion and Discussion. The data confirmed significant positive correlations between degree of KFC and presence of resistance to passive knee motion, cognitive impairment, impaired ambulation, and presence of knee pain. Physical therapy assessment and intervention may be appropriate in nonambulatory nursing home residents with resistance to passive motion, residents with KFC approaching 20 degrees, and ambulatory residents with minimal KFC who develop resistance and begin to regress in ambulation. Several areas for future study are suggested.