A new generation of antiepileptic drugs (AEDs) has reached the market in recent years with ten new compounds: felbamate, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, tiagabine, topiramate, vigabatrin and zonisamide. The newer AEDs in general have more predictable pharmacokinetics than older AEDs such as phenytoin, carbamazepine and valproic acid (valproate sodium), which have a pronounced inter-individual variability in their pharmacokinetics and a narrow therapeutic range. For these older drugs it has been common practice to adjust the dosage to achieve a serum drug concentration within a predefined 'therapeutic range', representing an interval where most patients are expected to show an optimal response. However, such ranges must be interpreted with caution, since many patients are optimally treated when they have serum concentrations below or above the suggested range. It is often said that there is less need for therapeutic drug monitoring (TDM) with the newer AEDs, although this is partially based on the lack of documented correlation between serum concentration and drug effects. Nevertheless, TDM may be useful despite the shortcomings of existing therapeutic ranges, by utilisation of the concept of 'individual reference concentrations' based on intra-individual comparisons of drug serum concentrations. With this concept, TDM may be indicated regardless of the existence or lack of a well-defined therapeutic range. The ten newer AEDs all have different pharmacological properties, and therefore, the usefulness of TDM for these drugs has to be assessed individually. For vigabatrin, a clear relationship between drug concentration and clinical effect cannot be expected because of its unique mode of action. Therefore, TDM of vigabatrin is mainly to check compliance. The mode of action of the other new AEDs would not preclude the applicability of TDM. For the prodrug oxcarbazepine, TDM is also useful, since the active metabolite licarbazepine is measured. For drugs that are eliminated renally completely unchanged (gabapentin, pregabalin and vigabatrin) or mainly unchanged (levetiracetam and topiramate), the pharmacokinetic variability is less pronounced and more predictable. However, the dose-dependent absorption of gabapentin increases its pharmacokinetic variability. Drug interactions can affect topiramate concentrations markedly, and individual factors such as age, pregnancy and renal function will contribute to the pharmacokinetic variability of all renally eliminated AEDs. For those of the newer AEDs that are metabolised (felbamate, lamotrigine, oxcarbazepine, tiagabine and zonisamide), pharmacokinetic variability is just as relevant as for many of the older AEDs. Therefore, TDM is likely to be useful in many clinical settings for the newer AEDs. The purpose of the present review is to discuss individually the potential value of TDM of these newer AEDs, with emphasis on pharmacokinetic variability.