1. In dogs anaesthetized with chloralose and artificially ventilated, the carotid sinuses were vascularly isolated and perfused with arterial blood. Mean aortic pressure was held constant at 111 +/- 2 mmHg (mean +/- S.E.M., n = 18) by means of a pressure bottle connected to the aorta and a Starling resistance. Both vagus nerves were sectioned in the neck and propranolol hydrochloride (1 mg kg-1 h-1) or atenolol (0.5-4 mg kg-1 h-1) was administered. The left and right renal blood flows were measured by electromagnetic flowmeters (wrap-round probes), glomerular filtration rate by creatinine clearance, urinary sodium by flame photometry and solute excretion by osmometry. 2. In six dogs decreasing pressure in the isolated carotid sinuses from 119 +/- 4 to 78 +/- 3 mmHg (n = 9) resulted in significant decreases in renal blood flow by 18 +/- 3% (P less than 0.01), glomerular filtration rate by 41 +/- 9% (P less than 0.01), filtration fraction by 30 +/- 11% (P less than 0.05), urine flow by 46 +/- 6% (P less than 0.001), sodium excretion by 46 +/- 9% (P less than 0.001) and osmolar excretion by 44 +/- 6% (P less than 0.001). Fractional sodium excretion did not change significantly. Increasing carotid sinus pressure back to 120 +/- 4 mmHg (n = 6) resulted in increases in all the variables to values not significantly different from those at initial high carotid pressure. 3. Ligation of left renal nerves at low carotid sinus pressure (83 +/- 3 mmHg, n = 5) caused significant increases in all of the variables in the left kidney. After ligation, changes in carotid sinus pressure produced no effect on the denervated left kidney, but in the three dogs in which the responses of the right kidney were also tested, the usual responses after denervation of the left kidney were seen. 4. The results show that changes in carotid sinus pressure around the normal range can result in significant reflex effects on renal haemodynamics and function and that these effects are mediated solely by renal sympathetic nerves. The influence of any extrarenal humoral factors seems to be minimal.