Polygraphic 16 h sleep recording were carried out in 35 adult cats following i.p. injections of various alpha-adrenoceptors agonists, antagonists and their combinations. The direct alpha-agonists, clonidine (CLO 0.005, 0.01 and 0.02 mg/kg) and xylazine (XYL 0.5, 1 and 2 mg/kg), dose-dependently decreased paradoxical sleep (PS) and deep slow wave sleep (S2), with a respective increase mainly in drowsy waking (D). alpha-Methyldopa, precursor of the potent alpha-agonist, alpha-m-noradrenaline (alpha-m-NA) suppressed PS, with little effect on other vigilance stages. Of the alpha-antagonists only phentolamine (PHE 10 and 20 mg/kg) increased significantly the 16 h mean of PS. Thymoxamine (THY 5 mg/kg) gave a modest, temporary increment in PS between 4 and 8 h after the injection, but the effect diminished with 10 mg/kg THY. Yohimbine (YOH 2 mg/kg) induced an early increment in aroused waking (A). Tolazoline (TOL 6 mg/kg) and THY (5 and 10 mg/kg) increased D in the first 4 h epoch. Phenoxybenzamine (PBZ 10 mg/kg) significantly decreased the 16 h mean of S2 and PS. PHE antagonized the PS suppressing effect of CLO (0.01 mg/kg) already at the dose of 5 mg/kg and with 10 and 20 mg/kg its PS increasing character prevailed. TOL (6 mg/kg) and YOH (2 mg/kg) were also effective antagonists to CLO. THY (5 and 10 mg/kg) was ineffective in this respect and clearly potentiated the S2 inhibiting effect of CLO. PBZ (10 mg/kg) powerfully potentiated both PS and S2 suppressing effects of CLO. PHE (20 mg/kg) was tested against XYL (0.5 and 1 mg/kg) and alpha-methyldopa (100 mg/kg). It also antagonized the PS inhibiting action of these drugs. All the three agonists preferentially stimulate presynaptic (alpha 2) type of alpha-adrenoceptors, inhibitory to noradrenaline (NA) transmission. Furthermore, as only antagonists possessing presynaptic potency inhibited PS suppression by alpha 2-agonists, while preferential alpha 1-antagonists were either ineffective or potentiated this effect, the results favor the hypothesis of a positive involvement of NA in the mechanisms of PS. The optimal level of NA transmission for PS may, however, be postulated to lie below that for arousal, in which case the balanced blockade of alpha 1- and alpha 2-adrenoceptors by PHE might be exceptionally favorable to PS. The possible role of alpha-adrenoceptive influences on cholinergic and 5-HT neurons and their relevance to alpha 2-agonist-induced sedation and inhibition of PS and S2 are discussed.