1. The potential influences of nitric oxide (NO) and prostaglandins on the renal effects of angiotensin II (Ang II) have been investigated in the captopril-treated anaesthetized rat by examining the effect of indomethacin or the NO synthase inhibitor, N(omega)-nitro-L-arginine methyl ester (L-NAME), on the renal responses obtained during infusion of Ang II directly into the renal circulation. 2. Intrarenal artery (i.r.a.) infusion of Ang II (1-30 ng kg(-1) min(-1)) elicited a dose-dependent decrease in renal vascular conductance (RVC; -38+/-3% at 30 ng kg(-1) min(-1); P < 0.01) and increase in filtration fraction (FF; +49+/-8%; P < 0.05) in the absence of any change in carotid mean arterial blood pressure (MBP). Urine output (Uv), absolute (UNaV) and fractional sodium excretion (FENa), and glomerular filtration rate (GFR) were unchanged during infusion of Ang II 1-30 ng kg(-1) min(-1) (+6+/-17%, +11+/-17%, +22+/-23%, and -5+/-9%, respectively, at 30 ng kg(-1) min(-1)). At higher doses, Ang II (100 and 300 ng kg(-1) min(-1)) induced further decreases in RVC, but with associated increases in MBP, Uv and UNaV. 3. Pretreatment with indomethacin (10 mg kg(-1) i.v.) had no significant effect on basal renal function, or on the Ang II-induced reduction in RVC (-25+/-7% vs -38+/-3% at Ang II 30 ng kg(-1) min(-1)). In the presence of indomethacin, Ang II tended to cause a dose-dependent decrease in GFR (-38+/-10% at 30 ng kg(-1) min(-1)); however, this effect was not statistically significant (P=0.078) when evaluated over the dose range of 1-30 ng kg(-1) min(-1), and was not accompanied by any significant changes in Uv, UNaV or FENa (-21+/-12%, -18+/-16% and +36+/-38%, respectively). 4. Pretreatment with L-NAME (10 microg kg(-1) min(-1) i.v.) tended to reduce basal RVC (control -11.8+/-1.4, +L-NAME -7.9+/-1.8 ml min(-1) mmHg(-1) x 10(-2)), and significantly increased basal FF (control +15.9+/-0.8, +L-NAME +31.0+/-3.7%). In the presence of L-NAME, renal vasoconstrictor responses to Ang II were not significantly modified (-38+/-3% vs -35+/-13% at 30 ng kg(-1) min(-1)), but Ang II now induced dose-dependent decreases in GFR, Uv and UNaV (-51+/-11%, -41+/-14% and -31+/-17%, respectively, at an infusion rate of Ang II, 30 ng kg(-1) min(-1)). When evaluated over the range of 1-30 ng kg(-1) min(-1), the effect of Ang II on GFR and Uv were statistically significant (P < 0.05), but on UNaV did not quite achieve statistical significance (P=0.066). However, there was no associated change in FENa observed, suggesting a non-tubular site of interaction between Ang II and NO. 5. In contrast to its effects after pretreatment with L-NAME alone, Ang II (1-30 ng kg(-1) min(-1)) failed to reduce renal vascular conductance in rats pretreated with the combination of L-NAME and the selective angiotensin AT1 receptor antagonist, GR117289 (1 mg kg(-1) i.v.). This suggests that the renal vascular effects of Ang II are mediated through AT1 receptors. Over the same dose range, Ang II also failed to significantly reduce GFR or Uv. 6. In conclusion, the renal haemodynamic effects of Ang II in the rat kidney appear to be modulated by cyclooxygenase-derived prostaglandins and NO. The precise site(s) of such an interaction cannot be determined from the present data, but the data suggest complex interactions at the level of the glomerulus.