Classical anesthetics of the γ-aminobutyric acid type A receptor (GABAA)-enhancing class (e.g., pentobarbital, chloral hydrate, muscimol, and ethanol) produce analgesia and unconsciousness (sedation). Dissociative anesthetics that antagonize the N-methyl-D-aspartate (NMDA) receptor (e.g., ketamine, MK-801, dextromethorphan, and phencyclidine) produce analgesia but do not induce complete loss of consciousness. To understand the mechanisms underlying loss of consciousness and analgesia induced by general anesthetics, we examined the patterns of expression of c-Fos protein in the brain and correlated these with physiological effects of systemically administering GABAergic agents and ketamine at dosages used clinically for anesthesia in rats. We found that GABAergic agents produced predominantly delta activity in the electroencephalogram (EEG) and sedation. In contrast, anesthetic doses of ketamine induced sedation, followed by active arousal behaviors, and produced a faster EEG in the theta range. Consistent with its behavioral effects, ketamine induced Fos expression in cholinergic, monoaminergic, and orexinergic arousal systems and completely suppressed Fos immunoreactivity in the sleep-promoting ventrolateral preoptic nucleus (VLPO). In contrast, GABAergic agents suppressed Fos in the same arousal-promoting systems but increased the number of Fos-immunoreactive neurons in the VLPO compared with waking control animals. All anesthetics tested induced Fos in the spinally projecting noradrenergic A5–7 groups. 6-hydroxydopamine lesions of the A5–7 groups or ibotenic acid lesions of the ventrolateral periaqueductal gray matter (vlPAG) attenuated antinociceptive responses to noxious thermal stimulation (tail-flick test) by both types of anesthetics. We hypothesize that neural substrates of sleep-wake behavior are engaged by low-dose sedative anesthetics and that the mesopontine descending noradrenergic cell groups contribute to the analgesic effects of both NMDA receptor antagonists and GABAA receptor-enhancing anesthetics.