Nitric oxide-stimulated modification of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) by [adenylate-32P]NAD has been interpreted in recent reports as ADP-ribosylation. Incubations of GAPDH with the NO-releasing agent sodium nitroprusside (SNP) and NAD resulted, however, in essentially equal incorporation of radiolabel from the adenine, phosphate, and nicotinamide moieties to the extent of approximately 0.02 mol of NAD.mol of GAPDH-1. Modification of GAPDH by free adenosine 5'-diphosphoribose (ADP-ribose) was only 10% of that by NAD. Exposure of GAPDH modified by NAD in the presence of SNP to HgCl2, which acts at thiol linkages, released two products. Both contained nicotinamide and adenylate but did not cochromatograph with NAD. GAPDH activity was inhibited by SNP in a dose-dependent manner in the presence of NAD. When inhibition was 80%, with 1 mM SNP and 1 mM dithiothreitol, covalent modification with NAD was < 2%. This result is consistent with the conclusion that inhibition of GAPDH activity by SNP in the presence of NAD is due primarily to active-site nitrosylation, as reported by other workers, and is not due to the minor modification with NAD. These results demonstrate that NO-stimulated modification of GAPDH with NAD is not ADP-ribosylation as previously reported but rather is covalent binding of NAD through a NO-dependent thiol intermediate, possibly providing an example of an unexpected, altered reactivity of a nitrosylated protein.