Kinetic assay of serum ethanol was investigated by predicting maximal product absorbance at 340 nm (A(mk)) through fitting to the yeast alcohol dehydrogenase reaction curve with the integrated rate equation, taking into account product inhibition in the presence of semicarbazide. Predicted A(mk) linearly responded to the preset constant of steady-state concentration of acetaldehyde (C(ald)). An exponential correlation function was established between desired C(ald) and putative A(mk) for authentic ethanol. For unknown samples, iterative fitting to reaction curve till preset constant C(ald) and resultant A(mk) satisfied this exponential correlation function yielded A(mk) with variation coefficient <4.3%. Variations in enzyme activity, data range and kinetic parameters showed negligible effects. The recovery was consistent to 100% with resistance to methanol and isopropanol. The upper limit of linear response for A(mk) was about 40 times of the lower limit. These results indicated that this kinetic method was reliable for serum ethanol assays with obvious advantages.