Approximately 30% of epileptic patients are resistant to monotherapy and may require combinations of antiepileptic drugs (AEDs). The combinations listed below have been evaluated in the maximal electroshock test in mice which is a model of human tonic-clonic and complex partial seizures. Gabapentin when combined with carbamazepine (CBZ), oxcarbazepine (OXC), lamotrigine (LTG), phenytoin (PHT), phenobarbital (PB), topiramate (TPM), or tiagabine (TGB) resulted in synergy with no neurotoxicity. However, LTG and PB produced elevations of the plasma concentrations of gabapentin. A synergy was also observed when LTG was combined with valproate (VPA) or TPM and these combinations were accompanied by antagonism concerning adverse effects. In contrast, LTG co-administered with CBZ produced a clear-cut anticonvulsant antagonism and neurotoxic additivity. Even worse results were found for OXC+LTG where anticonvulsant antagonism and neurotoxic synergy were evident. TGB combined with PHT, CBZ, PB, LTG or TPM led to additivity in the convulsive test and an apparent synergy of TGB+VPA was associated with a pharmacokinetic interaction. Additive effects were also observed for OXC+PB, CBZ, or VPA as regards anticonvulsant activity or neurotoxicity. In contrast, a combination of OXC+PHT was antagonistic in the seizure test and additive in neurotoxic tests, although OXC combined with TPM was clearly synergistic in terms of anticonvulsant activity and additive in terms of neurotoxicity. One of the newer AEDs, levetiracetam, produced anticonvulsant synergy when combined with CBZ, OXC and TPM. In conclusion, one can hardly predict an outcome of a drug combination basing upon the mechanisms of action of AEDs. Also, only benefi cial combinations of AEDs need to be clinically evaluated.