This prospective study investigated whether subjects' participation in a home-delivered meal program was associated with reduced length of hospitalization (LOH) and better nutritional status compared to controls matched for age (+/-5 years), gender, race and ICD9 Code. Thirty recipients of home-delivered meals and 30 non-recipients (ages 62-96 years) were recruited from three hospitals. Medical records and subject interviews were used to obtain lifestyle, dietary intake, functional abilities, laboratory values, and LOH data. Heights and weights were taken using a portable digital scale and stadiometer. T-tests revealed no significant differences between the two groups for demographic and lifestyle characteristics. Two highly significant differences (p = 0.005) emerged for LOH (8.6+/-4.3 days versus 15.8+/-13.8 days) and total lymphocyte count (TLC) (2058.9+/-1321.0m 3 versus 1675.9+/-1110.3m 3 ) for recipients and controls, respectively. Serum hematocrit varied significantly (p=0.02), but hemoglobin (p=0.29), serum albumin and total cholesterol (p=0.23) did not. Mean values for recipients were higher. Disease severity had no significant differences in number of medications (8.07 vs. 8.40; p=0.79), body mass index (25.2 vs. 24.9, p=0.84) and exercise level (p=0.16). Home-delivered meals programs appear beneficial in reducing hospital costs by shortening LOH and in supporting the immune system (reflected by TLC) for chronically ill elderly.