For almost 20 years, anticonvulsant monotherapy has been regarded as the ideal method of management of epilepsy. This line of thought, while true, has mutated into a general view of polypharmacy with anticonvulsants as a sign of therapeutic failure. However, even when the ‘right’ drug is given at the ‘right’ dose, studies have shown that in around 30% of patients, epilepsy will not fully respond to anticonvulsant monotherapy. For these patients, polypharmacy with anticonvulsants is an inevitability.There are good reasons why the established anticonvulsants should not be particularly well tolerated in combination. However, anticonvulsant polypharmacy has become so ‘undesirable’ that there has been a failure to search for firm treatment strategies for refractory epilepsy. Nevertheless, there are strong arguments to suggest that the newer anticonvulsants will prove to be better tolerated in combination than their established counterparts, particularly in view of their paucity of drug interactions, more predictable pharmacokinetics and narrower spectrum of action.This review calls for the development of a rational plan for the treatment of refractory epilepsy. Proper double-blind trials should be carried out to compare the efficacy and tolerability of anticonvulsant drug combinations. This is a huge task that could be rationalised by concentrating on combinations that have been reported to have particular benefit, or those with neurochemical properties that suggest an additive or even synergistic effect.