OBJECTIVE: This study sought to compare inpatient hospitalization practices for treating anorexia nervosa at two time periods and to assess whether differences in practices affected treatment outcome. METHOD: Hospital charts were examined for all patients with a diagnosis of anorexia nervosa and a body mass index ≤17.5 kg/m2 at admission who were admitted to a university hospital during 1975–1980 (early) or 1990–1995 (later). RESULTS: The "later cohort" had significantly higher admission and discharge body mass indexes, shorter lengths of stay, higher rates of weight gain, and more readmissions. Predictors of rehospitalization for the later cohort were increased length of stay, prior hospitalizations, and more rapid weight gain. CONCLUSIONS: This study suggests that the use of hospitalization to treat anorexia nervosa has changed over time and is associated with increased rates of readmission.