Intracarotid infusions of l-phenylephrine at 1.0 nmol·kg–1·min–1 or.10 nmol·kg–1·min–1 were accompanied by increases in salivary protein, urea, magnesium and bicarbonate, and by decreases in osmolality, hydrogen ion activity, sodium, potassium and chloride relative to cholinergically stimulated saliva. Intravenous infusions of phenylephrine at the same dose rates had much less effect on salivary composition with the differences between the routes of administration being greatest for the higher dose rate. Propranolol administered with phenylephrine via the carotid artery, at an antagonist:agonist ratio of 10:1, was much more effective in blocking the phenylephrine-induced changes in salivary composition than equimolar infusion of phentolamine with phenylephrine. Simultaneous intracarotid infusions of either a β1-antagonist (CGP20712A) or a β2-antagonist (ICI118551) with phenylephrine showed that ICI118551 was more potent than CGP20712A at preventing the changes in salivary composition associated with phenylephrine administration. It was concluded that α1-adrenoreceptors were not present in functionally significant numbers in the gland and that the effect of phenylephrine on the kangaroo mandibular was mediated by β-adrenoreceptors predominantly of the β2-subtype. As the phenylephrine dose rates in the kangaroos were comparable with those used to determine α-adrenergic responses of eutherian salivary glands and as both propranolol and phentolamine appeared to have minor β-sympathomimetic activity, at least one subtype of β-adrenoreceptors in macropods may not be identical to its eutherian counterpart.