In the last 20 years, several second-generation antiepileptic drugs (AEDs) have been marketed. These newer drugs are expensive and have no established superiority over the first-generation compounds in terms of efficacy when used as monotherapy. A systematic review of economic studies dealing with the newer AEDs has been performed to put these drugs in a wider perspective. A number of economic analysis studies of second-generation AEDs have examined these compounds as monotherapy or adjunctive therapy. Almost all monotherapy studies showed newer AEDs as having similar effectiveness but significantly higher acquisition costs than first-generation drugs. The evidence from adjunctive therapy studies was more conflicting. Lamotrigine appeared to be a cost-effective drug when higher thresholds were used, or when savings were defined by the cost of surgery. Levetiracetam also appeared to be cost effective when the costs of surgical investigation were discounted.In a decision model that included quantification of the uncertainty associated with the decision regarding the cost effectiveness of AEDs, second-generation drugs used as monotherapy for newly diagnosed partial epilepsy produced similar benefits but were more expensive than older drugs. The newer AEDs were more effective but more expensive than existing monotherapies in patients with refractory partial epilepsy, but may be cost effective at higher thresholds, and continue to be cost effective in patients responding to the assigned drug. In patients with newly diagnosed generalized epilepsy, valproate was more cost effective than lamotrigine.The results of current economic studies are difficult to assess and compare because of a number of methodological drawbacks. Future studies should be implemented using a standardized approach to define the costs and outcomes of representative cohorts of patients with newly diagnosed epilepsy recruited from different countries and followed prospectively.