Multifunctional catalysts with distinct functional components are known to have much improved selectivity. However, the well-known proximity-dependent selectivity observed in several high profile experiments is yet to be understood. Here, we reveal that such dependence is closely associated with the kinetics involved. Based on reaction-diffusion dynamics together with kinetic Monte Carlo simulation on a coarse-grained model, one famous example of bifunctional catalysis, namely the proximity-dependent selectivity from carbon dioxide to liquid fuels on a bifunctional catalyst composed of HZSM-5 and In2O3, has been systematically examined. It is found that the diffusion kinetics of the intermediate methanol generated on In2O3 plays a decisive role for the selectively. For different In2O3/HZSM-5 proximities, the local methanol concentration induce a shift of the dominant process for subsequent methanol-to-hydrocarbon reactions inside HZSM-5, resulting in a preferred reaction window to generate favorable liquid fuels with profound high selectivity. Our findings emphasize the importance of the largely overlooked kinetic in the design of multifunctional catalysts.