This paper reviews the lifetime contributions of the author to the field of sleep-wakefulness (S-W), reinterprets results of the early studies, and suggests new conclusions and perspectives. Long-term cats with mesencephalic transection show behavioral/polygraphic rapid eye movement sleep (REMS), including the typical oculo-pupillary behavior, even when the section is performed in kittens prior to S-W maturation. REMS can be induced as a reflex. Typical non-rapid eye movement S (NREMS) is absent and full W/arousal is present only after a precollicular section. The isolated forebrain (IF) rostral to the transection exhibits all features of W/arousal and NREMS [with electroencephalographic (EEG) spindles and delta waves], arousal to olfactory stimuli, and including the appropriate oculo-pupillary behaviors. These features also mature normally after neonatal transection. REMS is absent from the IF. After deprivation there is NREMS pressure and rebound in the IF, but the decerebrate cat only shows pressure for REMS. Most IF reactions to pharmacologic agents are within expectations, except for the tolerance/withdrawal effects of long-term morphine use which are absent. In contrast, these effects are supported by the brainstem (i.e. seen in the decerebrate cat). In cats with ablation of the telencephalon, or diencephalic cats, delta waves are absent in the thalamus. EEG thalamic spindle waves are seen triggering S for only 4-5 days after ablation. Therefore, true NREMS is absent in chronic diencephalic cats although pre- and postsomniac behaviors persist. These animals are hyperactive and show a pronounced, permanent insomnia; however, a low dose of barbiturate triggers a dramatic REMS/atypical NREMS rebound. Cats without the thalamus (athalamic cats), initially show a dissociation between behavioral hyperactivity/insomnia and the neocortical EEG, which for 15-20 days exhibits only delta and slower oscillations. Fast, low-voltage W rhythms appear later on, first during REMS, but spindle waves and S postures are absent from the start, such that these cats also display only atypical NREMS. Athalamic cats also show barbiturate-sensitive insomnia. Cats with ablation of the frontal cortices or the caudate nuclei remain permanently hyperactive. They also show a mild, but significant hyposomnia, which is permanent in afrontal cats, but lasts for about a month in acaudates. The polygraphic/behavioral features of their S-W states remain normal. We conclude and propose that: (a) the control of the S-W system is highly complex and distributed, but is organized hierarchically in a well-defined rostro-caudal manner; the rostral-most or highest level (telencephalon), is the most functionally complex/adaptative and regulates the lower levels; the diencephalic/basal forebrain, or middle level, has a pivotal role in inducing switching between S and W and in coordinating the lowest (brainstem) and highest levels; (b) W can occur independently in both the forebrain and brainstem, but true NREMS- and REMS-generating mechanisms exist exclusively in the forebrain and brainstem, respectively; (c) forebrain and brainstem S-W processes can operate independently from each other and are preprogrammed at birth; this helps understanding normal and abnormal polygraphic/behavioral dissociations in humans and normal dissociations/splitting in aquatic mammals; (d) NREMS homeostasis is present in the IF, but only REMS pressure after deprivation persists in the decerebrate cat; (e) the thalamus engages in both NREMS and W; (f) insomnia in diencephalic cats is the result of an imbalance between antagonistic W- and S-promoting cellular groups in the ventral brain (normally modulated by the telencephalon); (g) the EEG waves, which are signature for each S-W state, appear to truly drive the concomitant behaviors, e.g. a hypothetical human IF could alternate between behavioral NREMS and W/arousal/awareness; (h) a role for REMS is to keep the individual sleeping at the end of the self-limiting NREMS periods. The need for accelerating research on telencephaling NREMS periods. The need for accelerating research on telencephalic S-W processes and downstream control of the lower S-W system levels is emphasized.