Human serum albumin is categorized into human mercaptalbumin (HMA) and human non-mercaptalbumin (HNA), according to the redox state of the cysteine residue at position 34. The ratio of HMA to total albumin (%HMA) is a novel biomarker of oxidative stress as well as protein nutritional status, but measuring %HMA normally requires an expensive analyzer such as HPLC and LC-MS, and can hardly be conducted in many clinical sites. To address this issue, we aimed to develop a methodological basis for estimating %HMA without these analyzers. An analytical method was investigated consisting of three steps, i.e., 1) removal of HMA from serum or plasma by using a thiol-binding resin (i.e., thereby obtaining a HNA fraction), 2) determination of both total albumin and HNA concentrations by a colorimetric assay or ELISA, and 3) calculation of %HMA. Proof-of-concept experiments, using serum and plasma samples of 4 adult volunteers, showed that the estimated value of %HMA obtained by this analytical method was significantly correlated with the theoretical value of %HMA determined by HPLC. The subsequent validation experiment, using 86 serum samples of pregnant women in the Japanese participants of SMILE Iwamizawa, also confirmed the significant association between the estimated and theoretical values of %HMA. This analytical method can be a basis to determine %HMA without using HPLC or LC-MS, contributing to the universalization of %HMA measurement as a clinical test.