Objective: Disparities in epilepsy treatment are not uncommon; therefore, we examined population-based estimates of initial antiepileptic drugs (AEDs) in new-onset epilepsy among racial/ethnic minority groups of older US Medicare beneficiaries. Methods: We conducted retrospective analyses of 2008–2010 Medicare administrative claims for a 5% random sample of beneficiaries augmented for minority representation. New-onset epilepsy cases in 2009 had ≥1 International Classification of Diseases, Ninth Revision (ICD-9) 345.x or ≥2 ICD-9 780.3x, and ≥1 AED, AND no seizure/epilepsy claim codes or AEDs in preceding 365 days. We examined AED use and concordance with Quality Indicators of Epilepsy Treatment (QUIET) 6 (monotherapy as initial treatment = ≥30 day first prescription with no other concomitant AEDs), and prompt AED treatment (first AED within 30 days of diagnosis). Logistic regression examined likelihood of prompt treatment by demographic (race/ethnicity, gender, age), clinical (number of comorbid conditions, neurology care, index event occurring in the emergency room (ER)), and economic (Part D coverage phase, eligibility for Part D Low Income Subsidy [LIS], and ZIP code level poverty) factors. Results: Over 1 year of follow-up, 79.6% of 3,706 new epilepsy cases had one AED only (77.89% of whites vs. 89% of American Indian/Alaska Native [AI/AN]). Levetiracetam was the most commonly prescribed AED (45.5%: from 24.6% AI/AN to 55.0% whites). The second most common was phenytoin (30.6%: from 18.8% Asians to 43.1% AI/AN). QUIET 6 concordance was 94.7% (93.9% for whites to 97.3% of AI/AN). Only 50% received prompt AED therapy (49.6% whites to 53.9% AI/AN). Race/ethnicity was not significantly associated with AED patterns, monotherapy use, or prompt treatment. Significance: Monotherapy is common across all racial/ethnic groups of older adults with new-onset epilepsy, older AEDs are commonly prescribed, and treatment is frequently delayed. Further studies on reasons for treatment delays are warranted. Interventions should be developed and tested to develop paradigms that lead to better care.