The rates of reaction of the "alpha" nucleophile, 2,3-butanedione monoximate (Ox(-)), and the "normal" nucleophile, 4-chlorophenoxide (4-ClPhO(-)), with the phosphorus ester, p-nitrophenyl diphenylphosphinate (PNPDPP), were measured in DMSO-H2O mixtures at 25.0 degrees C. Addition of small quantities of DMSO (up to ca. 10 mol %) to the aqueous solvent led to a decrease in the second-order rate constants of reaction for both nucleophiles. Additional : increases in the DMSO content above 20 mol % resulted in smooth and regular increases in the rate for 4-ClPhO(-), but for Ox(-) the corresponding increase in reaction rate upon increase of DMSO content is not linear, being best described as either curved or composed of two intersecting linear portions in the log k vs mole percent DMSO plot. The magnitude of the alpha-effect, k(Ox)-/k(4-ClPhO-), shows a bell-shaped dependence on solvent composition, with a maximum value of ca. 40 at ca. 50 mol % DMSO. The data is analyzed in terms of the recently developed novel procedure for Bronsted-type plots, in which changes in basicity and reactivity of the nucleophile are brought about through changes in solvent composition rather than changes of remote substituent. This analysis suggests that the bell-shaped dependence of the alpha-effect magnitude upon solvent composition may stem from the effect of advanced desolvation of the Ox(-) nucleophile at the transition state relative to-bond formation, i.e., imbalance or nonperfect synchronization effects.