Renal plasma clearances (C) of 14C-tetraethylammonium (TEA) and p-aminohippurate (PAH) as estimates of arterial renal plasma flow (ARPF) were evaluated in anesthetized rats during control conditions and during intravenous glucose infusion. Venous renal blood flow was measured directly by means of a servo-controlled pump, keeping the renal venous pressure constant. Arteriovenous extraction fractions (E = 1 - P(renal venous)/P(renal arterial)) for PAH averaged 88.3 +/- (SE) 0.8% in control rats and 82.0 +/- 0.9% in glucose-infused rats (p less than 0.001); E(TEA) averaged 92.0 +/- 0.6 and 90.1 +/- 0.6%, respectively (p less than 0.05). Under both experimental conditions, (C/E)PAH did not differ significantly from ARPF, while (C/E)TEA underestimated ARPF; the rate of extraction of TEA exceeded the rate of excretion by 15-20%, probably due to accumulation of TEA in renal tissue. It is concluded that, when corrected for E, C(PAH) is in general a more accurate estimate for ARPF than C(TEA). However, under conditions involving changes in plasma glucose levels C(TEA) may provide a better estimate of the effective renal plasma flow than C(PAH).