Aldose reductase, purified to homogeneity from bovine kidney, is converted in a temperature-dependent process from a low-Km/low-Vmax form to a high-Km/high-Vmax form of the enzyme. Activation, which results in significant changes in the protein secondary structure, as detected by fluorescence spectroscopy, circular dichroism, and thiol modification with 5,5'-dithiobis(2-nitrobenzoic acid), has no effect on the apparent Mr, pI, or homogeneity of the enzyme, as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and agarose isoelectric focusing. Vmax, which varied less than 3-fold for a series of aldehyde substrates with either activation state of the enzyme, increased an average of (17 +/- 4)-fold upon activation of the enzyme. V/Kaldehyde increased or decreased up to 4-fold, depending on the substrate. Activation desensitized the enzyme to inhibition by aldose reductase inhibitors, with the apparent Ki value increasing from 2-fold for Epalrestat [ONO-2235, (E)-3-(carboxymethyl)-(E)-5-[2-methyl-3-phenylpropenylidene]-rhoda nine] to 200-fold for AL-1576 (spiro [2,7-difluorofluorene-9,4'-imidazolidine]-2',5'-dione). Biphasic double-reciprocal plots for the aldehyde substrates and biphasic Dixon plots for inhibition by AL-1576 and Statil [ICI-128,436; 3-[(4-bromo-2-fluorobenzyl)-4-oxo-3H-phthalazin-l-ylacetic acid], observed during the course of activation, are quantitatively accounted for by the individual contributions of the two enzyme forms. On the basis of an analysis of the kinetic data, a mechanism is proposed in which isomerization of the free enzyme limits the rate of the forward reaction for the unactivated enzyme and is the primary step affected by activation.