In this study, the sulfhydryl (SH) contents of unheated and heated (90 °C, 5 min) soy protein were detected under different conditions (pH, reagent addition order, SDS/GuHCl concentration, EDTA) using two aromatic disulfide reagents: 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) and 4,4'-dithiodipyridine (DPS). Two fluorescent alkylating reagents, monobromobimane (mBBr) and N-(1-pyrenyl)maleimide (NPM), were chosen due to their high sensitivity and were also used. Amino acid analysis was used to detect the SH (cysteine) contents of unheated (7.51 ± 0.45 μmol SH/g protein) and heated (1.47 ± 0.10 μmol SH/g protein) soy protein, and similar results were obtained using enzymatic hydrolysis-assisted DPS. The SH content detected by DTNB was affected by pH, denaturant species, and denaturant concentration, and the best results were obtained at pH 7.0 when 6 M GuHCl was added after DTNB. These results were lower than that of the amino acid analysis, however. The SH detected by DPS was not as affected as that of DTNB by pH, denaturant species, and denaturant concentration. Additionally, the results of the amino acid analysis were similar to that of DPS at pH 7.0 in 2% SDS and 4-6 M GuHCl when SDS and GuHCl were added after DPS. EDTA did not have a significant effect on SH detection when DTNB and DPS were added before SDS and GuHCl. Finally, although mBBr and NPM can detect SH in low protein concentrations ((1)/10 of that required for DTNB and DPS), mBBr and NPM overestimated the SH content of soy protein. Therefore, using DPS at pH 7.0 when it is added before SDS and GuHCl is the most reliable method for detecting the SH content of soy protein.