Objective: This study aimed at estimating the cumulative incidence of antiepileptic drug (AED) treatment failure of first-line monotherapy levetiracetam vs valproic acid in glioma patients with epilepsy. Methods: In this retrospective observational study, a competing risks model was used to estimate the cumulative incidence of treatment failure, from AED treatment initiation, for the two AEDs with death as a competing event. Patients were matched on baseline covariates potentially related to treatment assignment and outcomes of interest according to the nearest neighbor propensity score matching technique. Maximum duration of follow-up was 36 months. Results: In total, 776 patients using levetiracetam and 659 using valproic acid were identified. Matching resulted in two equal groups of 429 patients, with similar covariate distribution. The cumulative incidence of treatment failure for any reason was significantly lower for levetiracetam compared to valproic acid (12 months: 33% [95% confidence interval (CI) 29%–38%] vs 50% [95% CI 45%–55%]; P <.001). When looking at specific reasons of treatment failure, treatment failure due to uncontrolled seizures was significantly lower for levetiracetam compared to valproic acid (12 months: 16% [95% CI 12%–19%] vs 28% [95% CI 23%–32%]; P < 0.001), but no differences were found for treatment failure due to adverse effects (12 months: 14% [95% CI 11%–18%] vs 15% [95% CI 11%–18%]; P =.636). Significance: Our results suggest that levetiracetam may have favorable efficacy compared to valproic acid, whereas level of toxicity seems similar. Therefore, levetiracetam seems to be the preferred choice for first-line AED treatment in patients with glioma.