1. The effects of exogenous NO and endothelial-derived NO (EDNO) on the afferent arteriole were investigated in the in vitro perfused hydronephrotic rat kidney. Vessels were pre-constricted with angiotensin II (0.1-0.3 nM) or KCl (30 mM). NO was infused directly into the renal artery at concentrations ranging from 30-9000 nM. ODQ (10, 30 microM) was administered to examine the effects of guanylyl cyclase inhibition. Kidneys were treated with ibuprofen (10 microM) to avoid actions of prostaglandins. 2. During angiotensin II-induced vasoconstriction, NO elicited vasodilation at concentrations of 30 900 nM (EC50=200 nM) and ODQ caused a 10 fold shift in NO-sensitivity (EC50 1600 nM). During KCl-induced vasoconstriction, NO elicited a maximal dilation of 82+9% at 9000 nM (EC50 2000 nM) and ODQ had no effect. Thus in the presence of ODQ, the NO concentration-response curves for KCI- and angiotensin II-induced vasoconstriction were identical (P>0.2). 3. To assess the possible role of cyclic GMP-independent mechanisms in the actions of EDNO, we compared the effects of L-NAME, ODQ and ODQ+L-NAME on acetylcholine-induced vasodilation. Angiotensin II reduced afferent arteriolar diameters from 16.7+/-0.5 to 8.1+/-0.8 microns and acetylcholine fully reversed this effect (16.9+/-0.5 microns). ODQ restored the angiotensin II response in the presence of acetylcholine (7.1+/-0.6 microns) and the subsequent addition of L-NAME had no further effect (6.8+/-0.7 microns). Similarly, L-NAME alone, fully reversed the actions of acetylcholine. 4. Our findings indicate that exogenous NO is capable of eliciting renal afferent arteriolar vasodilation through both cyclic GMP-dependent and cyclic GMP-independent mechanisms. The cyclic GMP-independent action of NO did not require K+ channel activation, as it could be elicited in the presence of 30 mM KCl. Finally, although cyclic GMP-independent effects of exogenous NO could be demonstrated in our model, EDNO appears to act exclusively through cyclic GMP.