Based on the reviewed literature and the data presented in this paper, conclusions can be drawn with respect to the validity of the WAG/Rij strain of rats as a model for absence epilepsy in humans. The view that the WAG/Rij model has "face validity" is supported by the simultaneous presence of clinical and electroencephalographic signs characterizing absences in rat and humans, by the decrease in responsiveness during the presence of spike-wave discharges in both species, by the agreement between model and patient with respect to the preferential occurrences of spike-wave discharges at transitions in states of vigilance, by the corresponding modulation of spike-wave discharges by physical and mental activities in both and, finally, by the fact that in both humans and rats absence epilepsy is inherited. Against this view, however, argue two points. In rats, absences appear after puberty and are maintained during life, while in humans the seizures occur before puberty and then disappear or convert to more serious forms of epilepsy. The second point is the frequency difference of the spikes and waves in the discharge train: 8-10 Hz in the rat and 3 Hz in the human (though there are no a priori reasons why the frequency of spike waves in the burst must be the same in all species). The absence model also has predictive validity, based on pharmacological data that demonstrate the specificity of certain drugs as being effective in convulsive epilepsies and not in absence epilepsy. So far, all drugs affect spike-wave activity the same way in rats and humans, with lamotrigine being, perhaps, the only exception. Furthermore, sleep deprivation is a powerful provocation for the initiation of spike-wave discharges in both rats and humans. Potential explanations for the presence of absence seizures in rats have been found at the levels of activities in networks and nuclei; of neurons, membrane properties, and ion channels; of proteins and enzymes; and, finally, of genes and chromosomes. Further descriptions of the cellular processes can be found extensively in the literature (e.g., McCormick and Contreras, 2001) and those of the thalamo-cortico-thalamic network in this review as well as in others (Avanzini et al., 1999). Considering the extensive involvement of the phenomena under study with theoretical issues such as the relationship between sleep spindles and spike-wave discharges, and the origin of seizure activity, it can be concluded that the model also has construct validity as far as the present neurobiological theories holding for absence epilepsy in humans are concerned. The WAG/Rij model can therefore be recommended for continued use in evaluating antiepileptic drugs for monotherapy and polytherapy, as well as for the toxicological side effects of putative new antiabsence drugs.