Although it has been widely recognized that elderly nursing home residents with dementia display a wide range of disruptive behaviors, little data are available on the course of those behaviors over time or effective methods to prevent or manage them. The overall goal of this study, therefore, is to test intervention strategies designed to reduce disruptive behaviors in nursing home residents with dementia. A secondary goal is to describe, over time, the occurrence and course of disruptive behaviors. Based on a series of planned preliminary studies and pilot data analysis, this study will compare the effectiveness of a Psychosocial Activity Intervention (PSA), an Activities of Daily Living Intervention (ADL), and a combination of the two interventions (CB) in decreasing the frequency, duration, and intensity of disruptive behaviors in demented elderly residents. The relationship between disruptive behaviors and psychological, cognitive, and functional status of the elders will be profiled. Data collected will ascertain the course and occurrence of disruptive behaviors in these elders. A controlled field experiment of the interventions will be tested with randomly assigned groups of a total of 100 residents (25 in each of four nursing homes) with a diagnosis of dementia who exhibit disruptive behaviors. Following pretesting, using a variety of behavioral, cognitive, and physical functioning status instruments, random assignments will be made to five groups: PSA, ADL, CB, a PSA Placebo group, or a No- Intervention Control Group. Each resident (with the exception of the control group) will follow a treatment structure and format. Post-testing will be conducted at one and two months after the treatment period. During the treatment period, residents will be videotaped during the treatment activity, AM care, and the noon meal. A longitudinal descriptive profile of all nursing home residents in two additional homes will be conducted with an anticipated total sample size of 150 (75 in each home). Using a battery of cognitive, psychological, observational, and physical examination measures, the elders' disruptive behaviors will be examined over two month periods for two years. Data analysis will consist of descriptive and correlational procedures. Comparison of the treatment groups will be analyzed using analysis of variance with preplanned contrasts. Longitudinal data will be analyzed using repeated measures and multivariate linear modeling techniques.