Epilepsy is one of the most common neurological disorders. Even though existing antiepileptic drugs can render 80% of newly diagnosed patients seizure free, a significant number of patients have chronic intractable epilepsy causing disability with considerable socioeconomic implications. There is, therefore, a need for more potent and effective antiepileptic drugs and drugs with fewer adverse effects, particularly CNS effects. Drugs for the treatment of partial seizures are particularly needed. With major advances in our understanding of the basic neuropathology, neuropharmacology and neurophysiology of epilepsy, numerous candidate novel antiepileptic drugs have been developed in recent years. This review comparatively evaluates the pharmacokinetics, efficacy and adverse effects of 12 new antiepileptic drugs namely vigabatrin, lamotrigine, gabapentin, oxcarbazepine, felbamate, tiagabine, eterobarb, zonisamide, remacemide, stiripentol, topiramate and levetiracetam (ucb-L059). Of the 12 drugs, vigabatrin, lamotrigine and gabapentin have recently been marketed in the UK. Five of these new drugs have known mechanisms of action (vigabatrin, lamotrigine, tiagabine, oxcarbazepine and eterobarb), which may provide for a more rational approach to the treatment of epilepsy. Oxcarbazepine, remacemide and eterobarb are prodrugs. Vigabatrin, gabapentin and topiramate are more promising on the basis of their pharmacokinetic characteristics in that they are excreted mainly unchanged in urine and not susceptible to significant pharmacokinetic interactions. In contrast, lamotrigine, felbamate and stiripentol exhibit significant drug interactions. Essentially, all the drugs are effective in partial or secondarily generalised seizures and are effective to varying degrees in other seizure types. Particularly welcome is the possible effectiveness of zonisamide in myoclonus and felbamate in Lennox-Gastaut syndrome. In relation to adverse effects, CNS effects are observed with all drugs, however, gabapentin, remacemide and levetiracetam appear to exhibit least. There is also the possibility of rational duotherapy, using drugs with known mechanisms of action, as an additional therapeutic approach. The efficacy of these 12 antiepileptic drug occurs despite the fact that candidate antiepileptic drugs are evaluated under highly unfavourable conditions, namely as add-on therapy in patients refractory to drug management and with high seizure frequency. Thus, whilst candidate drugs which do become licensed are an advance in that they are effective and/or are associated with less adverse effects than currently available antiepileptic drugs in these patients, it is possible that these drugs may exhibit even more improved risk-benefit ratios when used in normal clinical practice.