Treatment-resistant seizures affect about a third of patients suffering from epilepsy. To fulfill the need for new medications targeting treatment-resistant seizures, a number of rodent models offer the opportunity to assess a variety of potential treatment approaches. The use of such models, however, has proven to be timeconsuming and labor-intensive. In this study, we performed pharmacological characterization of the allylglycine (AG) seizure model, a simple in vivo model for which we demonstrated a high level of treatment resistance. (D,L)-Allylglycine inhibits glutamic acid decarboxylase (GAD) -the key enzyme in γ-aminobutyric acid (GABA) biosynthesis -leading to GABA depletion, seizures, and neuronal damage. We performed a side-by-side comparison of mouse and zebrafish acute AG treatments including biochemical, electrographic, and behavioral assessments. Interestingly, seizure progression rate and GABA depletion kinetics were comparable in both species. Five mechanistically diverse antiepileptic drugs (AEDs) were used. Three out of the five AEDs (levetiracetam, phenytoin, and topiramate) showed only a limited protective effect (mainly mortality delay) at doses close to the TD 50 (dose inducing motor impairment in 50% of animals) in mice. The two remaining AEDs (diazepam and sodium valproate) displayed protective activity against AG-induced seizures. Experiments performed in zebrafish larvae revealed behavioral AED activity profiles highly analogous to those obtained in mice. Having demonstrated crossspecies similarities and limited efficacy of tested AEDs, we propose the use of AG in zebrafish as a convenient and high-throughput model of treatment-resistant seizures.