The asymmetric hydrogenation of methyl pyruvate to methyl lactate, by cinchonidium functionalized MCM-41 supported [Ru4(μ-H) 3(CO)12]- as the precatalyst has been studied kinetically and by scanning transmission electron microscopy (STEM). Existence of an induction time and two competitive equilibriums are inferred from the time monitored conversion data. Steady state approximation gives a poor fit, but a kinetic model (Eley-Rideal) consisting of a fast equilibrium between methyl pyruvate and the catalyst, a slow one between the catalyst and dihydrogen, and a rate determining reaction between the latter and methyl pyruvate, accurately simulates the time monitored conversion profiles. The model suggests that on increasing the methyl pyruvate concentration there is a change in the stoichiometry of the equilibrium between the catalyst and the methyl pyruvate. The change in enantioselectivity with time can also be accurately simulated by assuming enantiomeric excess to be proportional to the rate constant for methyl lactate formation. Both kinetic and STEM data strongly suggest that in the fresh catalyst the bare metal cluster framework is retained, and under the catalytic conditions agglomeration of the subnano sized clusters leading to the formation of nanoparticles of ruthenium is a relatively slow process. A hypothetical enantionface selection mechanism consistent with the empirical rate law, previous reports, and other experimental evidence is proposed. © 2011 American Chemical Society.