Despite lack of evidence for the efficacy of physical restraints with the elderly, little systematic research and no clinical trials in nursing homes have tested interventions aimed at decreasing their use. Therefore, the purpose of this prospective, controlled study is to: 1) Investigate the relative effects of two experimental interventions, restraint education and restraint education-with- consultation, on the use of physical restraints in nursing homes; 2) Examine the relative effects of these interventions on resident outcomes (functional, cognitive, affective, health, and care plans) and staff outcomes (burnout, employment stability, beliefs about restraint efficacy and knowledge of alternatives); and 3) Test the relationship between nursing hours or serious resident injuries and restraint use. A quasi-experimental, potentially nonequivalent control group design will be employed to compare three nursing homes which have been randomly assigned to one of three treatments. A minimum of 375 residents over age 60 and 156 clinical staff in three nursing homes will comprise the two subsamples. Intervention I, Restraint Education, consists of 10, 45-60 minute programs taught over a six month period to increase staff knowledge regarding physical restraint use and comprehensive assessment and management of problematic behaviors. Intervention II, Restraint Education-with-Consultation, consists of an identical program of Restraint Education supplemented by 12 hours per week of direct consultation to staff. A master's prepared gerontologic nurse specialist will implement the two experimental interventions. Intervention III, Control, consists of neither Restraint Education nor Restraint Education-with Consultation. Resident and staff subjects will be measured in two preliminary observations spaced three months apart; again immediately prior to and immediately following the interventions; and at 3 months and 6 months post-intervention to observe both immediate and sustained intervention effects. The design for hypothesis testing includes repeated measures, cross-sectional, and correlational analyses.