We used a hybrid quantum-mechanics/molecular-mechanics (QM/MM) approach to simulate the adsorption of Au-n (n = 1-5), AuPd, and Au2Pd2 clusters inside the TS-1 and S-1 pores. We studied nondefect and metal-vacancy defect sites in TS-1 and S-1 for a total of four different environments around the T6 crystallographic site. We predict stronger binding of all clusters near Ti sites in Ti-substituted framework compared to adsorption near Si sites-consistent with the experimental finding of a direct correlation between the Ti-loading and the Au-loading on the Au/TS-1 catalysts with high Si/Ti ratio. The cluster binding is also stronger near lattice-metal vacancies compared to fully coordinated, nondefect sites. In all the cases, a trend of binding energy (BE) versus Au cluster size (n) shows a peak at around n = 3-4. Our results show that there is enough room for the attack of H2O2 on the Ti-defect site even with Au1-4 adsorbed-a result that supports the possibility of H2O2 spillover from the Au clusters to the adjacent Ti-defect sites. Mulliken charge analysis indicates that in all the cases there is electron density transfer to adsorbed clusters from the zeolite lattice. In the case of both gas-phase and adsorbed Au-Pd clusters, all the Pd atoms were positively charged, and all the Au atoms were negatively charged due to the higher electron-affinity of Au. We also found a correlation between the BE and the charge transfer to the clusters ( the higher the charge transfer to the clusters, the higher the BE), and a universal correlation was found for Au2-5 when BE and charge transfer were plotted on a per atom basis. A relatively larger charge transfer to the adsorbed clusters was found for the Ti sites versus the Si sites, and for the defect sites versus the nondefect sites. The trends in the BE were corroborated using Gibbs free energy of adsorption (Delta G(ads)), and the implications of Delta G(ads) in sintering of Au clusters are also discussed. Our results confirm that electronic factors such as cluster-charging are potentially important support effects for the Au/TS-1 catalyst.