Background & Purpose: Drug-resistant epilepsy affects 30% of patients who continue to have seizures despite current antiseizure medications (ASMs). Preclinical drug screening uses acute dosing and evoked seizures, which may not fully capture the complexities of drug resistance and human treatment regimens. We introduce a novel experimental paradigm that incorporates clinical treatment regimens, pharmacokinetic monitoring, and behavioral tolerability assessments to accurately model drug-resistant epilepsy in animal models. Experimental Approach: Rats with epilepsy following kainic acid induced status epilepticus were enrolled in a triple cross-over study to evaluate the dose-dependent efficacy and tolerability of three commonly used ASMs — lamotrigine, levetiracetam, and carbamazepine. Each medication was delivered in rodent chow for two weeks via our automated system, maintaining steady-state exposures measured by pharmacokinetic sampling. Seizure control was monitored via 24/7 videoEEG, and behavioral tolerability was evaluated using minimal motor impairment and hyperexcitability assays. Key Results: Chronic oral dosing with carbamazepine and levetiracetam reduced seizure frequency by more than 50% in over half the animals at clinically relevant doses. Lamotrigine, however, was either ineffective or worsened seizures at toxic doses, increasing both convulsive and clustered seizures. Levetiracetam was well tolerated, while carbamazepine impaired motor function at the highest dose. Lamotrigine led to increased hyperactivity and aggressive behavior at all doses. Conclusions: This study highlights the need for preclinical models that better reflect human epilepsy, considering both efficacy and side effects in drug development. Our findings emphasize the complexity of drug responses and underscore the importance of improved models for drug development.