Although anticonvulsant polytherapy has been widely and traditionally used in the treatment of epilepsy, there is little evidence of its advantages over monotherapy. It does, however, lead to problems of chronic toxicity, drug interactions, failure to evaluate individual drugs, and sometimes exacerbation of seizures. There are many causes of polytherapy which could be avoided by more careful monitoring and supervision of therapy. Studies in new, previously untreated referrals suggest there is considerable potential for monotherapy. In the event of failure of optimum monotherapy, the value of polytherapy is not yet clear. In chronic patients on polytherapy there may be scope for careful rationalization to two or sometimes one drug, with reduction in chronic toxicity and sometimes improved seizure control. Reduction of therapy, however, may be impossible or hazardous due to withdrawal seizures. Even after successful reduction, seizure control is much less satisfactory than in new referrals. It is easier to avoid polytherapy than to reduce it. There is a need to define more carefully the limits of effective anticonvulsant therapy.