1. We tested the renal effects of the alpha 2-adrenoceptor agonists, rilmenidine and guanabenz and the antagonists, 2-methoxyidazoxan and idazoxan, in conscious dogs. Our aim was to test the hypothesis that putative imidazoline (I) receptors influence renal function. We reasoned that since rilmenidine and guanabenz are selective for I1- and I2-binding sites respectively, an influence of one of these receptive sites on renal function would be reflected in qualitative differences between the effects of these agents. Moreover, effects mediated by putative I-receptors should be relatively resistant to antagonism by the selective alpha 2-adrenoceptor antagonist, 2-methoxyidazoxan. Since the effects of these drugs on renal function could be mediated in the central nervous system or periphery, the dogs were studied under both normal and ganglion-blocked conditions. 2. In dogs with intact autonomic reflexes, 2-methoxyidazoxan (15 micrograms kg-1 plus 0.6 micrograms kg-1 min-1) produced effects consistent with a generalized increase in sympathetic drive, including increases in mean arterial pressure and plasma renin activity, and a reduction in sodium excretion. In ganglion-blocked dogs, 2-methoxyidazoxan reduced sodium excretion but had no discernible effect on systemic or renal haemodynamics. We conclude that an alpha 2-adrenoceptor-mediated mechanism in the central nervous system tonically inhibits sympathetic drive in the conscious dog. 3. In ganglion-blocked dogs idazoxan (3-300 micrograms kg-1) dose-dependently increased arterial pressure. This was not abolished by concomitant administration of 2-methoxyidazoxan (0.3-30 micrograms kg-1). The pressor effect of idazoxan is therefore probably mediated by an agonist action at alpha 1-adrenoceptors. 4. The effects of infusions of rilmenidine (0.1-1.0 mg kg-1) and guanabenz (10-100 micrograms kg-1) were indistinguishable. They comprised dose-dependent increases in mean arterial pressure, urine excretion, and glomerular filtration rate (the latter in ganglion blocked dogs only), and dose-dependent reductions in heart rate, renal blood flow and sodium excretion (only in dogs with intact autonomic reflexes). All of these effects were antagonized by 2-methoxyidazoxan. 5. We conclude that the renal effects of rilmenidine and guanabenz infusions in conscious dogs are predominantly, if not completely, attributable to activation of alpha 2-adrenoceptors. Our results do not support the hypothesis that putative I-receptors contribute towards the renal effects of these agents.