The renal excretory responses to volume expansion (VE), by 10 % body wt, were determined in groups of anaesthetised streptozotocin-induced diabetic rats with one denervated and one innervated kidney in the presence and absence of nitric oxide synthase (NOS) inhibitors. VE in diabetic rats increased (P < 0.001) cumulative urine sodium excretion (CuUNaV) to 104 ± 9 and 69 ± 6 µmol min-1 (g kidney wt)-1 in the denervated and in the innervated kidneys, respectively, which were both less (P < 0.001) than in the non-diabetic rats, at 225 ± 14 and 148 ± 14 µmol min-1 (g kidney wt)-1, respectively, in the denervated and the innervated kidney. The non-selective NOS inhibitor, NG-nitro-L-arginine-methyl-ester (L-NAME) given to the diabetic rats with intact renal innervation enhanced CuUNaV after VE by 43 % (P < 0.001), while the combination of L-NAME and renal denervation restored CuUNaV to a value comparable to that of non-diabetic rats. In diabetic rats treated with either a relatively selective inhibitor for the neuronal isoform of NOS, 7-nitroindazole, or a relatively selective inhibitor for the inducible isoform of NOS, aminoguanidine, CuUNaV after VE was similar to the untreated diabetic rats irrespective of whether or not the renal nerves were present. This investigation demonstrated that NO production contributed, at least partly, to the depressed ability to excrete a saline load in diabetes mellitus. The endothelial isoform of NOS was most probably responsible for generating NO which caused the blunted excretory responses. The ability of NO to attenuate the excretory responses to volume expansion was an action independent of the renal innervation status. Experimental Physiology (2001) 86.4, 481-488.