SUMMARY1. Stimulation of the renal sympathetic nerves in pentobarbitone anaesthetized rats achieved a 13 % reduction in renal blood flow, did not change glomerular filtration rate, but reduced urine flow by 37 %, absolute sodium excretion by 37 %, and fractional sodium excretion by 34 %. Following inhibition of converting enzyme with captopril (0-38 mmol kg-' h-1), similar nerve stimulation reduced both renal blood flow and glomerular filtration rate by 16 %, and although urine flow and absolute sodium excretion fell by 32 and 31 %, respectively, the 18 % fall in fractional sodium excretion was significantly less than that observed in the absence of captopril.2. Renal nerve stimulation at low levels, which did not change either renal blood flow or glomerular filtration rate, reduced urine flow, and absolute and fractional sodium excretions by 25, 26 and 23 %, respectively.3. In animals receiving captopril at 0-38 mmol kg-' h-1, low-level nerve stimulation caused small increases in glomerular filtration rate of 7 % and urine flow of 12 %, but did not change either absolute or fractional sodium excretions. At one-fifth the dose of captopril (0-076 mmol kg-' h-'), low-level nerve stimulation did not change renal haemodynamics but decreased urine flow, and absolute and fractional sodium excretions by 10, 10 and 8 %, respectively. 4. These results showed that angiotensin II production was necessary for regulation of glomerular filtration rate in the face of modest neurally induced reductions in renal blood flow and was compatible with an intra-renal site of action of angiotensin II preferentially at the efferent arteriole. They also demonstrated that in the rat the action of the renal nerves to decrease sodium excretion was dependent on angiotensin II.