Brønsted coefficients have been determined for the rates of thiol-disulfide interchange between low molecular weight thiols and the disulfide groups of four native or modified proteins: DNase (beta nuc congruent to 0.36), lysozyme (beta nuc congruent to 0.55), adenylate kinase(SSCH3)2 (beta nuc congruent to 0.65), and papain(SSCH3) (beta nuc congruent to 0.45). These values are similar to those observed for reductions of oxidized glutathione and Ellman's reagent by a similar set of thiols (beta nuc congruent to 0.5). Glutathione is anomalously slow in reduction of certain protein disulfide groups: although this effect may in part reflect steric hinderance to attack by the relatively large glutathione molecule at disulfides shielded by protein tertiary structure, other (presently undefined) factors appear also to be important, at least in the case of DNase. The rates of reduction of several disulfide derivatives of papain(SSR) by DTT were determined. These data provide estimates of the Brønsted coefficient for the "central" thiol in thiol-disulfide interchange: these estimates fall in the range beta c congruent to -0.25 to -0.43. Rates of reduction of protein disulfide moieties were analyzed by using a Brønsted equation developed previously [Szajewski, R. P., & Whitesides, G. M. (1980) J. Am. Chem. Soc. 102, 2011] to yield pKa values for the participating thiol moieties: in particular, for papain, pKa(Cys-25) = 8.4 at pH 9 and pKa (Cys-25) = 4.1 at pH 6. The thiols of the structurally essential cysteine group of lysozyme seem to have pKa congruent to 11. The advantages and disadvantages of this method for estimating thiol pKa values are discussed.