Recombinant human brain calbindin D(28K) (rHCaBP), human Cu,Zn-superoxide dismutase (HCuZnSOD), rabbit muscle glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and bovine serum albumin (BSA) were found to be S-glutathiolated in decomposed S-nitrosoglutathione (GSNO) solutions. Tryptic or Glu-C digestion and MALDI-TOF MS analyses of the digests are consistent with S-thiolation of Cys111 and Cys187 of HCuZnSOD and rHCaBP, respectively, upon exposure to decomposed GSNO. GAPDH activity analysis reveals that S-glutathiolation most likely occurs on the active site Cys149, and the single free Cys34 is assumed to be the site of S-glutathiolation in BSA. The yields of S-glutathiolation of rHCaBP, GAPDH, and BSA were much higher than those of HCuZnSOD. The latter is limited by the accessibility of Cys111 to the glutathiolating reagent in the HCuZnSOD dimer. Unlike decomposed GSNO, fresh GSNO, reduced glutathione (GSH), and oxidized glutathione (GSSG) are not efficient S-glutathiolating agents for the proteins examined here. On the basis of analysis by mass spectrometry and UV-visible absorption, GSNO decomposition in the dark at room temperature yields glutathione disulfide S-oxide [GS(O)SG], glutathione disulfide S-dioxide (GSO(2)SG), and GSSG as products. GS(O)SG is the efficient protein S-glutathiolating agent in GSNO solutions, not GSNO, which does not carry out efficient S-glutathiolation of rHCaBP, HCuZnSOD, or GAPDH in vitro. A hydrolysis pathway yielding GSOH and nitroxyl (HNO/NO(-)) as intermediates is proposed for GSNO decomposition in the dark. This is based on inhibition of GSNO breakdown by dimedone, a reagent specific for sulfenic acids, and on nitroxyl scavenging by metmyoglobin. The results presented here are contrary to numerous reports of protein S-thiolation by low-molecular weight S-nitrosothiols.