OBJECTIVES: To validate the recently updated Charlson Comorbidity Index (CCI) for the prediction of healthcare utilization and to compare its predictive power in patients with diabetes or asthma. METHODS: Data were retrieved from the Medical Expenditure Panel Survey (MEPS) Panel 12 (2007-2008) for this retrospective cohort study. The original CCI (CCI-1) and updated CCI (CCI-2) scores were calculated for patients who had diabetes or asthma in 2007. Adjusted R2 from linear regression models were used for the estimation of log-transformed healthcare expenditures (COST) in 2008. C statistics from logistic regression models were used to compare the predictive power of the risk of hospitalizations  (≥ 1 admission), risk of emergency department visits (≥ 1 visit), and high expenditures (≥ 90th percentile of COST) in 2008. RESULTS: 833 diabetic patients and 704 asthmatic patients were included in the study. The diabetes cohort had a mean age of 59.7 years (SD: 15.3), and 54% were female; the asthma cohort had a mean age of 37.8 years (SD: 23.6), and 59% were female. In the linear regressions, the CCI-2 explained more variance in COST in diabetic patients than in asthmatic patients (adjusted R2 = 17.4% vs. 14.1%), adjusting for demographics. The CCI-2 was a better predictor of high COST (c = 0.881 vs. 0.816) and the risk of hospitalization (c = 0.713 vs. 0.682) but a poorer predictor of the risk for an emergency department visit (c = 0.583 vs. 0.653) in the diabetes cohort than in the asthma cohort. In both cohorts, the CCI-2 (c = 0.583 to 0.881) exhibited consistently better predictive power than the CCI-1 (c = 0.576 to 0.839). CONCLUSIONS: The predictive power of CCI varies depending on the outcomes of interest in patients with diabetes or asthma. The updated CCI showed improved predictive performance compared to the original CCI.