Nursing home residents typically have decreased functional and physical status and high health care utilization and costs. This randomized trial evaluates whether physical therapy is beneficial for frail debilitated long-stay residents of nursing homes. Subjects are recruited from a cohort of academic and community nursing home residents who have resided in the nursing home for greater than 3 months and are over age 60 and dependent in at least two activities of daily living. Subjects randomized to the intervention group receive one-on-one physical therapy sessions three times weekly for 4 months, while control group subjects receive structured social visits three times weekly to control for potential Hawthorne effects. Physical therapy sessions generally last 30 minutes and consist of functional activity and general conditioning exercises; these exercises are individually tailored to the subject's level of physical and functional disability. Prime outcome variables are physical function assessed by an observer-administered, performance-based instrument and self-perceived health status assessed by the Sickness Impact Profile. Health care utilization and associated costs are calculated for the following areas: the nursing home, hospitalizations, outpatient visits and procedures, medications, and the intervention. A cost-effectiveness ratio dividing incremental health care utilization and physical therapy intervention costs by the observed improvement in physical function is calculated. It is expected that results of this study can be used to help determine whether long-stay nursing home residents should be eligible for physical therapy.