BACKGROUND: Loss of ambulatory ability with acute hospitalization is common and often does not improve by discharge. OBJECTIVES: To define admission predictors of regaining ambulatory ability during hospitalization in patients with expected activity limitations. DESIGN: Prospective cohort study. SETTING: University teaching hospital. PARTICIPANTS: Two hundred and eighty-six patients at least 55 years of age whose activity was expected to be limited to a bed or chair for at least the first 5 days of hospitalization or who had a hip fracture, who were ambulatory in the 4 weeks prior to hospital admission, and whose length of stay in the hospital was less than 32 days. MEASUREMENTS: Baseline data collected from admission physician and nurse interviews and abstracted from the medical records included length of stay, demographic characteristics, global health measures, presence of specific diseases, and hospital-related factors hypothesized to affect ambulation. Nurses were asked weekly if patient activity was still expected to be limited to a bed or chair. RESULTS: Despite initially being limited to a bed or chair, 42% had regained ambulatory ability by discharge. Recovery of ambulatory ability was independently associated with not being married (odds ratio [OR] = 3.0, 95% confidence interval [CI] 1.4-6.2), higher physician-rated life expectancy (OR = 1.9, 95% CI 1.3-2.8), absence of restraints (OR = 2.5, 95% CI 1.2-5.5), having a urinary catheter (OR = 2.2, 95% CI 1.2-5.5), having deep vein thrombosis (OR = 11.4, 95% CI 1.2-105.1), and having a higher level of bed mobility at admission (OR = 1.7, 95% CI 1.1-2.6). CONCLUSIONS: Recovery of ambulatory ability is closely associated with physician-rated life expectancy and hospital-related factors, particularly those that affect mobility. Early recognition of who will recover ambulatory ability may help with discharge planning and potential interventions. (c) 2006 Society of Hospital Medicine.