Recently a new mechanism for the enantioselective hydrogenation of (activated) ketones has been suggested, involving a zwitterionic intermediate between the tertiary amine function of the chiral modifier and the keto-carbonyl group of the reactant. The present NMR study indicates that the mechanistic model is probably based on erroneous interpretation of the experimental data; the NMR spectra that have been reported for zwitterion formation may arise from an aldol addition product between the ketone and solvent acetone. Steric effects and the regioselectivity of the hydrogenolysis of the hypothetic zwitterionic intermediate also exclude this mechanism for ketone hydrogenation on Pt. (c) 2005 Elsevier Inc. All rights reserved.