Excess NO generation plays a major role in the hypotension and systemic vasodilatation characteristic of sepsis. Yet the kidney response to sepsis is characterized by vasoconstriction resulting in renal dysfunction. We have examined the roles of inducible nitric oxide synthase (iNOS) and endothelial NOS (eNOS) on the renal effects of lipopolysaccharide administration by comparing the effects of specific iNOS inhibition, -N6-(1-iminoethyl)lysine (L-NIL), and 2,4-diamino6-hydroxy-pyrimidine vs. nonspecific NOS inhibitors (nitro- -arginine-methylester). cGMP responses to carbamylcholine (CCh) (stimulated, basal) and sodium nitroprusside in isolated glomeruli were used as indices of eNOS and guanylate cyclase (GC) activity, respectively. LPS significantly decreased blood pressure and GFR (112+/-4 vs. 83+/-4 mmHg; 2.66+/-0.29 vs. 0. 96+/-0.22 ml/min, P < 0.05) and inhibited the cGMP response to CCh. GC activity was reciprocally increased. L-NIL and 2, 4-diamino-6-hydroxy-pyrimidine administration prevented the decrease in GFR (2.71+/-0.28 and 3.16+/-0.18 ml/min, respectively), restored the normal response to CCh, and GC activity was normalized. In vitro application of L-NIL also restored CCh responses in LPS glomeruli. Neuronal NOS inhibitors verified that CCh responses reflected eNOS activity. L-NAME, a nonspecific inhibitor, worsened GFR (0.41+/-0.15 ml/min), a reduction that was functional and not related to glomerular thrombosis, and eliminated the CCh response. No differences were observed in eNOS mRNA expression among the experimental groups. Selective iNOS inhibition prevents reductions in GFR, whereas nonselective inhibition of NOS further decreases GFR. These findings suggest that the decrease in GFR after LPS is due to local inhibition of eNOS by iNOS, possibly via NO autoinhibition.