Baroreceptor responsiveness was investigated in conscious dogs following increasing doses (i.v.) of the selective alpha-adrenoceptor agonists methoxamine (alpha 1) and oxymetazoline (alpha 2), in the presence and absence of beta-adrenoceptor blockade. The study was repeated in another group of dogs with background afferent baroreceptor nerve activity reduced by continuous infusion of sodium nitroprusside. Both agonists dose dependently increased mean arterial pressure and reflexly decreased heart rate. In dogs pretreated with a beta-adrenoceptor antagonist a correlation between increase in mean arterial pressure (increase up to 70 mmHg) and increase in heart period (baroreceptor responsiveness) indicated no difference in the regression lines between methoxamine and oxymetazoline for both the normotensive and the sodium nitroprusside groups. However, in the dogs not pretreated with a beta-adrenoceptor antagonist the slope of the regression line for oxymetazoline was steeper than that for methoxamine (P less than 0.01) in the normotensive group. In the sodium nitroprusside group the regression line for oxymetazoline was situated significantly to the left of the methoxamine line (P less than 0.05). It is suggested that this greater bradycardic response to the alpha 2-adrenoceptor agonist oxymetazoline was caused by suppression of the cardiac sympathetic component (presynaptic modulation of noradrenaline release) in addition to the vagal activation and the sympathetic withdrawal component of the reflex. This indicates that drugs with alpha 2-adrenoceptor agonistic activity can influence a reflex physiological situation under conditions of low sympathetic nerve activity.