1. In two groups of Saffan-anaesthetized rats, we studied the role of the renin-angiotensin system in mediating the antidiuresis and antinatriuresis induced by moderate systemic hypoxia. 2. In both groups, a first period of hypoxia (breathing 12% O2 for 20 min) induced a fall in arterial partial pressure of O2 (Pa,O2; to 42 mmHg), a fall in mean arterial pressure (MABP), no change in renal blood flow (RBF) due to an increase in renal vascular conductance (RVC = RBF/MABP) and falls in urine flow and absolute sodium excretion (UNaV). Concomitantly, plasma renin activity increased from 3.08 +/- 0.68 (mean +/- S.E.M.) to 8.36 +/- 1.8 ng ml-1 hr-1. 3. In group 1 (n = 11), Losartan (10 mg kg-1, I.V.), the angiotensin (AII) AT1 receptor antagonist, induced a fall in MABP (115 +/- 3 to 90 +/- 3 mmHg), an increase in RVC such that RBF was unchanged, and falls in glomerular filtration rate (GFR), urine flow and UNaV. However, hypoxia induced qualitatively similar changes to those seen before Losartan treatment. 4. In group 2 (n = 9), we occluded the aorta distal to the renal artery to prevent basal MABP and renal perfusion pressure (RPP) from falling after addition of Losartan and to keep the hypoxia-induced fall in MABP the same as before Losartan treatment. Nevertheless, Losartan induced an increase in basal RVC, RBF, urine flow and UNaV whilst hypoxia induced falls in urine flow and UNaV that were proportionately similar to those seen prior to addition of Losartan. 5. These results indicate that in the Saffan-anaesthetized rat, AII exerts tonic, renal vasoconstrictor and consequent antidiuretic and antinatriuretic influences in normoxia, but does not contribute to the hypoxia-induced antidiuresis and antinatriuresis. We propose that renin secretion is increased by the hypoxia-induced fall in RPP rather than by an increase in renal sympathetic activity. Thus, the AII generated cannot produce antidiuresis and antinatriuresis by its known facilitatory influence on the actions of an increase in sympathetic activity on the renal tubules and is insufficient to produce these effects by direct actions. Rather, these results support the view that the antidiuresis and antinatriuresis of moderate hypoxia is predominantly due to the fall in RPP.