Milstein, and co-workers reported that the complex [(bpy)Rh(hd)]+PF6- (bpy = substituted bipyridine ligand, hd = 1,5-hexadiene) shows catalytic activity in the hydrogenation of acetone (Tollner, K. et al. Science 1997, 278, 2100). The activity in an ordered monolayer was found to be dramatically greater than in solution. We used the DFT functional mPW1K (Lynch, B. J. et al. J. Phys. Chem. A 2000, 104, 4811) to investigate the mechanism of the homogenous reaction. The suitability of the mPW1K functional was verified by coupled cluster calculations on a model system. Bulk solvent effects were considered. Various alternative catalytic cycles were evaluated, and we found that one potential mechanism involves metal-catalyzed keto-enol tautomerization to form [(bpy)Rh(enol)](+) that adds hydrogen yielding a complex with axial and equatorial hydride ligands. The reaction continues via transfer of the hydrides to the enolic C=C bond thereby forming 2-propanol and regenerating the catalyst. Another potential catalytic cycle involves formation of [(bpy)Rh(acetone)(2)(H)(2)](+), which has a spectator solvent ligand, and initial transfer of the equatorial hydride to the carbonyl carbon of acetone. Other mechanisms involving hydrogen transfer to the acetone tautomer involved higher barriers. With an eye toward modeling multi-center catalysis, various model systems for the bpy ligand were considered. It was found that diimine (HN=CH-CH=NH) compares very well with bpy, whereas cis-1,2-diiminoethylene (H2C=N-CH=CH-N=CH2) yields a reaction profile very close to that of bpy. Finally, the system with two rhodium centers, [(diimine)Rh](2)(2+), was investigated. The results strongly suggest that an enol-type catalytic cycle occurs and that cooperativity between the two metal centers is responsible for the acceleration of the reaction in the monolayer system.