1929 The effect of a 24 week home-based program on functional performance was examined in 72 frail men and women (mean age 77±5; range 70-92). Subjects (57 women and 15 men) were randomly assigned to either a nutrition education attention control group (NUT: n=38) or a progressive strength and balance training program (EX; n=34). Subjects in EX group received progressive hand and ankle weights and limited supervision (11 home visits). They were expected to exercise 3 days per week on their own and were encouraged to increase their leisure time activities. Subjects in NUT received limited supervision (8 home visits) and were instructed to eat at least 5 fruits and vegetables and 3 calcium rich foods each day. There was an average of 83% compliance to both interventions. All assessments were performed by an investigator blinded to subject group assignment. A Short Physical Performance Battery test (including tandem stand, gait speed and chair stand time) improved by 26.2±37.5% in EX and decreased by 1.2±22.1% in NUT (P=0.001). Functional capacity as measured by the Physical Performance Test increased by 6.1±13.4% in EX and decreased by 2.8±13.6% in NUT (P=0.02). Dynamic balance (20 foot tandem walk in seconds) improved by 33.8±14.4% in EX and by only 11.5±23.7% in NUT (P=0.0002). There was no difference between groups in change in chair stand time, muscle strength (1 repetition max) or distance covered in six minute walk. There was no relationship between change in these variables or dynamic balance with functional performance. In conclusion: 1) functional performance can be improved by a moderate home-based exercise program with limited supervision; 2) it is possible that some non-physiologic effect of the exercise intervention was etiologic in the change in functional performance observed. Supported by: the Brookdale Foundation, National Institute of Aging, US Department of Agriculture, Tufts University and the Tufts Associated Health Plan