AbstractObjective: To test the a priori hypothesis that obesity is a predictor of risk for reporting homebound status.Research Methods and Procedures: A longitudinal cohort study was conducted with 21, 645 community‐dwelling men and women 65 to 97 years old. A nutrition risk screen was administered baseline between 1994 and 1999 and again 3 to 4 years later. Univariate analyses identified baseline variables associated with subsequent reporting of homebound status. Multivariable logistic regression models were created to identify baseline variables that were significant independent predictors of reporting homebound status.Results: At baseline, 24% of the cohort had BMI ≥ 30. There were 12, 834 (45% men) respondents at follow‐up (68% response). Non‐responders at follow‐up differed little from responders except for greater baseline age (72.2 ± 6.2 vs. 71.4 ± 5.6 years, p < 0.001) and reporting of any functional limitations (9.2% vs. 4.9%, p < 0.001). At follow‐up, those who reported homebound status (n = 169) were significantly (p < 0.001) older (80.3 ± 7.3 vs. 75.1 ± 5.5 years) and more likely to report functional limitations (83.4% vs. 10.8%). Univariate analyses identified 16 baseline variables that were eliminated stepwise until five significant independent predictors remained: age ≥ 75 years (2.21, 1.55 to 3.15/odds ratio, 95% confidence interval), BMI ≥ 35 (1.75, 1.04 to 2.96), poor appetite (2.50, 1.29 to 4.86), low income (1.59, 1.00 to 2.56), and any functional limitation (10.67, 7.36 to 15.46).Discussion: Obesity remained a significant independent predictor for reporting homebound status and should be considered in screening of older populations and in the planning, implementation, and evaluation of services for homebound older persons.