The intent of this study is to determine the effects of a mild to moderate exercise program (tailored for ambulatory, semi ambulatory and non ambulatory elderly subjects) on joint motion, muscle size and strength, body composition/anthropometric measurements, cardiac performance, and morale. The sample will consist of elderly persons living in congregate housing, facilitated housing, and long term care institutions including nursing homes. These populations have not been adequately represented in exercise studies to date. Previous studies have focused primarily on highly independent healthy elders. The sample will consist of 180 subjects with 90 in the experimental and 90 in the control group. The exercise program is tailored to persons who are ambulatory (on floor-mat exercises), semi ambulatory (chair exercises), and non ambulatory (in-bed exercises). Comparable muscle groups/joints are involved in the three exercise formats. The design is a 3x2 time series factoral design. Data analysis will consist of t tests, ANOVA and MANOVA for time series data.