The discovery of genetically determined epileptic syndromes associated with specific mutations of genes codifying for subunits of voltage or ligand-activated ion channels highlights the role of ion channels in epileptogenesis. In vitro and in vivo models of channel pathology have been used to define the functional consequence of the mutations identified in human epilepsies. The evaluation of gene-channel mutations based on molecular and physiological techniques have provided significant knowledge on the cellular mechanisms leading to inherited human epilepsies, and possibly to nongenetic human epilepsies due to "acquired" channel pathologies. We review some of the studies that have explored human epileptic disorders through experimental manipulations of these channels, highlighting some of the difficulties that have arisen using "in vitro" preparations or rodent models. These findings underscore the need for further studies to address the mechanisms involved in mutated-channel dysfunctions.