OBJECTIVES: To assess and compare the predictive and discriminative performances of comorbidity indexes for health care outcomes and evaluate comorbidity dimensionality using psychometric techniques. METHODS: The sample was type 2 diabetes in the Medicaid setting from 2003 to 2007. The conceptual framework was based on the Aday-Anderson’s Healthcare Utilization model. Four comorbidity indexes targeted were the Charlson Comorbidity Index, Elixhauser Index (EI), Chronic Disease Score (CDS), and Health related Quality of Life Comorbidity Index (HRQL-CI). Three types of outcomes were health care behaviors, including physician treatment adherence and patient medication adherence, utilization and expenditures. Multiple regression analyses assessed the predictive performance of comorbidity index. The c statistic (the area under the receiver operator curve) evaluated discriminative validity of the comorbidity index. Confirmatory factor analysis identified comorbidity dimensionality. The SAS™, STATA™, and LISREL™ statistical software were utilized. RESULTS: A total of 9832 patients were finally included, with mean age of approximate 45 years and the majority of them was female (73%) and White (52%). The CDS demonstrated the best performance in predicting physician treatment adherence and discriminating medication adherence behavior. The CDS and HRQL-CI mental aspect index had better predictive validity for medication adherence and similar discrimination for physician treatment adherence. Diagnosis-driven indexes (e.g., EI) had better performances for health care utilization and expenditures outcomes compared to medication-based index (CDS). A 7-factor pattern/dimensionality was noticed and it provided best model fit and predictive performance across different health care outcomes. Individual comorbidity dimensions demonstrated differential impacts for a given outcome. CONCLUSIONS: The CDS and HRQL-CI mental aspect index served as better risk adjustment tools for studying healthcare behaviors. Diagnosis-driven indexes remained the first choice for health care utilization and expenditures data. Comorbidity index which accounts for comorbidity dimensionality provided better risk adjustment and insightful knowledge regarding the impacts of different features of comorbidities in predicting patient outcomes.