OBJECTIVES: The purpose of the study was to compare pharmacy-based and diagnosis-based indexes in predicting the total healthcare expenditure in current year and next year for the adult asthmatic population.  METHODS: Data from panel 11 involving asthmatic patients aged 18 years or more were selected from the 2006-2007 Medical Expenditure Panel Survey (MEPS) The diagnosis-based index was coded according to the D’Hoore version of the Charlson comorbidity index, and the pharmacy-based index was defined using Chronic Disease Score-1 (CDS-1). The performance of both indexes was evaluated for  the current year and next year. Additional variables included in the multivariable model were demographic characteristics, the number of prescribed medications and refills (NPM), and the number of emergency visits.  RESULTS: The mean age (± Standard Deviation) in the study was 47.48±0.87 years and the mean total health expenditure was $8,285±500. Most of the selected patients were female (68%) and the vast majority of them were white (82%). The CDS-1 (adj R2 for current year=0.2483, and adj R2 for next year = 0.2469) performed better than D’Hoore (adj R2 for current year=0.1286, and adj R2 for next year = 0.1271) when they were added individually. The best predictive model included age, sex, NPM, and either D’Hoore (adj R2 for current year = 0.4334 and adj R2 for next year =0.4294) or CDS-1 (adj R2 for current year =0.4346 and adj R2 for next year =0.4304).  CONCLUSIONS: The best predictive model for the current and the prospective year was the model that includes demographics, number of prescriptions, and either pharmacy-based or diagnosis-based indexes