A new hypothesis is offered regarding the pathomechanism of generalized epilepsy with spike-wave paroxysms (GESw) based on the pertaining literature and personal investigations. The first section is devoted to a critical overview of the development of theories regarding GESw. The centrencephalic theory, the debate on subcortical versus cortical origin, the "corticoreticular" hypothesis of Gloor and, finally, the "dyshormic" concept of Niedermeyer are outlined. In the next section it is shown that there is a particular optimum zone between sleep and wakefulness and between REM and slow wave sleep which highly favours the occurrence of spike-wave paroxysms. According to our investigations into the dynamics within this critical zone, the spike-wave paroxysms always appear with characteristic fluctuations of the level of consciousness where the changes towards awakening are always followed by rebounds towards sleep. Hence, the dynamic properties of this unstable border zone become especially interesting in the genesis of spike-wave paroxysms. It has been shown that even without epilepsy, a dynamics can be observed in the micro-oscillations in the depth of sleep which could be interpreted according to the reciprocal induction regulation model. In our concept the process of falling asleep emerges from rebounds of the sleep promoting system in response to sensory inputs streaming in from the external environment. According to this model, arousal influences in sleep have a sleep promoting effect. We interpret in this way all synchronized EEG reactions elicited by sensory stimuli and we consider K-complex type synchronization reactions as a "building stone" of the process of falling asleep which contains the whole process in concentrated form. The manifold similarities between the K-complex and the spike-wave pattern are demonstrated. On this basis spike-wave paroxysms can be regarded as an epileptic "caricature" of the sleep induction momentum reflected in the K-complex phenomenon. Hence, the GESw is the epileptic disorder of the sleep promotion function. This hypothesis resolves and explains many contradictory features of our knowledge about this mechanism and gives a new biologically oriented framework for further research. In the light of the hypothesis it has been attempted to interpret some of the characteristic features of the GESw: the genetic determination, the age dependency, the link with the sleep-waking cycle as well as the functional-anatomical characteristics and the symptoms of the seizures.