OBJECTIVES:  To examine relationships between different comorbidity indices and healthcare costs, and to compare the performance of these indices in predicting healthcare costs among elderly gynecologic cancer survivors. METHODS:  This retrospective study used 2007-2010 Surveillance, Epidemiology, and End Results (SEER)-Medicare data. The primary independent variables were comorbidity indices [Charlson Comorbidity Index (CCI), Elixhauser Comorbidity Index (EI), National Cancer Institute (NCI) index, Chronic Disease Score (CDS) and RxRisk]. The dependent variables were costs for emergency room (ER)/inpatient visits, outpatient visits, office-based practitioner visits, prescriptions, and total healthcare. Gamma regressions with a log link were used for the analyses. The Akaike Information Criterion, Bayesian Information Criterion, and Likelihood ratio tests were used to compare the predictive ability of the indices. RESULTS:  The mean total annual healthcare cost for the 4,063 survivors studied was $40,605 (SD=$34,014). In the unadjusted models, indices (except the CDS-2 and RxRisk) were positively significantly associated with ER/inpatient visit and total healthcare costs. In the adjusted models, CCI and CDS-1 scores and several EI indicators were positively significantly associated with higher ER/inpatient visit and total healthcare costs after controlling for covariates. None of the indices (except a few EI indicators) were associated with outpatient visit, office-based practitioner visit, or prescription costs in the models. CCI and CDS-1 scores outperformed other indices in predicting ER/inpatient visit and total healthcare costs. CONCLUSIONS:  This study found meaningful associations between different comorbidity indices and ER/inpatient visit and total healthcare costs among elderly gynecologic cancer survivors in the US. Study findings suggest using the CCI and CDS-1 scores in predicting ER/inpatient visit and total healthcare costs in this population. This study may help payers in their budgeting by identifying comorbid conditions associated with higher costs, and health economists choose the better comorbidity indices to use in their research.