OBJECTIVES:  It is a common practice in observational studies using administrative claims data to control for confounding using comorbidity scores. The Charlson Comorbidity Index (CCI) is perhaps one of the most widely used measures. However, the CCI was developed for predicting inpatient deaths. The Centers for Medicare and Medicaid Services Chronic Condition Warehouse (CCW) algorithms for identifying 27 chronic conditions and the RxRisk prescription-based comorbidity score offer alternative measures that may be more suitable for studies of outpatient utilization. This study examined the ability of all three comorbidity measures to predict outpatient healthcare utilization. METHODS: A retrospective analysis was conducted using Mississippi Medicaid administrative claims data for all beneficiaries continuously enrolled from January 1, 2014 to December 31, 2016. For each beneficiary, the first outpatient visit in 2015 was considered as the index date. CCI, RxRisk, and CCW scores were calculated for each beneficiary from medical and pharmacy data for the 12-month period prior to the index date. Outpatient healthcare utilization was calculated as the number of outpatient care visits occurring during the 12 months post index date. Three separate prediction models were run for each of the three comorbidity scoring techniques. All three models had gender, age, and race as covariates. Model performance was assessed using the Rfor the models. RESULTS:  The model with using CMS CCW measure had an R of 17.6% when predicting outpatient healthcare utilization. This model had better predictive power than the CCI model (R = 14.6%) and the RxRisk model (R= 15.6%). CONCLUSIONS:  Both the CMS CCW and the RxRisk comorbidity measures were better at predicting outpatient healthcare utilization. When conducting observational studies on outpatient healthcare events researchers should consider using comorbidity measures that better capture chronic conditions that predict outpatient healthcare utilization rather than the CCI measure.