The definition of drug-resistant epilepsy (DRE) is elusive and still controversial owing to some unresolved questions such as: how many drugs should be tried before a patient is considered intractable; to which extent side-effects may be acceptable; how many years are necessary before establishing drug resistance. In some cases, the view of epilepsy as a progressive disorder constitutes another important issue. Despite the use of new antiepileptic drugs (AEDs), intractable epilepsy represents about 20-30% of all cases, probably due to the multiple pathogenetic mechanisms underlying refractoriness. Several risk factors for pharmacoresistance are well known, even if the list of clinical features and biological factors currently accepted to be associated with difficult-to-treat epilepsy is presumably incomplete and, perhaps, disputable. For some of these factors, the biological basis may be common, mainly represented by mesial temporal sclerosis or by the presence of focal lesions. In other cases, microdysgenesis or dysplastic cortex, with abnormalities in the morphology and distribution of local-circuit (inhibitory) neurons, may be responsible for the severity of seizures. The possible influence of genes in conditioning inadequate intraparenchimal drug concentration, and the role of some cytokines determining an increase in intracellular calcium levels or an excessive growth of distrophic neurites, constitute other possible mechanisms of resistance. Several hypotheses on the mechanisms involved in the generation of DRE have been indicated: (a) ontogenic abnormalities in brain maturation; (b) epilepsy-induced alterations in network, neuronal, and glial properties in seizure-prone regions such as the hippocampus; (c) kindling phenomenon; (d) reorganization of cortical tissue in response to seizure-induced disturbances in oxygen supply. Such hypotheses need to be confirmed with suitable experimental models of intractable epilepsy that are specifically dedicated, which have until now been lacking.