The last decade has seen the licensing of nine new antiepileptic drugs (AEDs) with more to come. Despite this, only 58 and 63% of patients with localisation-related and newly diagnosed epilepsy, respectively, had been seizure-free for more than a year in separate prospective outcome studies undertaken at the Epilepsy Unit in Glasgow. Data will be presented to support the hypothesis that adolescent and adult epileptic patients comprise two distinct populations. Around 60% will be controlled on monotherapy with the first or second choice AED, while the majority of the remainder is difficult-to-control. It is for this latter group and the many pharmacoresistant paediatric patients with encephalopathic syndromes that we need new AEDs. For a successful clinical outcome, patients must be able to tolerate the treatment. Neurotoxic, sedative, cognitive and psychiatric symptoms, dysmorphic and other long-term side effects, and teratogenesis plague the current crop of AEDs. Pharmacokinetic and pharmacodynamic interactions complicate the situation still further. These problems may, in part, be a consequence of combining drugs with similar mechanisms of action. Unravelling the genetics of the epilepsies will provide a range of tempting targets for pharmacological intervention. We need, also, models of refractory epilepsy to help identify promising therapies. An efficient regulatory trial programme will ensure rapid availability of new AEDs for the many children and adults whose lives continue to be blighted by seizures.