Zeolite-supported transition metal catalysts, which couple the unique size- and shape-selectivity arising from the well-defined microporous structure of zeolites with the inherent high activity of metal species, have demonstrated remarkable performance in numerous catalytic reactions. Conventionally, such catalysts are prepared by loading metal species in the micropores of zeolites in the form of clusters (each containing only several atoms). Despite their high catalytic activity, the ultra-small clusters are usually highly mobile, and tend to migrate from the micropores to the crystal surfaces of zeolite during the reaction, where they agglomerate and deactivate. In this dissertation, we attempted to solve this issue by encapsulating metal nanoparticles (NPs) in zeolite crystals, based on the following considerations: (i) compared to clusters, nanoparticles have similar catalytic activity but much less mobility; and (ii) as long as the active sites are inside the zeolite crystals (not necessarily in the micropores that are too small to accommodate nanoparticles), they can exhibit selectivity associated with the zeolite structure. In the first chapter, we gave a general introduction to zeolites and zeolite supported catalysts, focusing on the preparation of hierarchical zeolites that are the main catalyst support materials used in the research projects of this dissertation. In the second chapter, we encapsulated highly dispersed Pd NPs (~2.6 nm) in zeolite ZSM-5 crystals, and used the obtained catalyst (Pd@SG-ZSM-5) for the hydrogenation of cinnamaldehyde. The confinement effect gave rise to an interesting catalytic behavior: compared with the traditional supported Pd catalyst prepared by impregnation, Pd@SG-ZSM-5 showed a 2.5-fold enhancement in the selectivity of hydrocinnamaldehyde (73% vs. 30%). Liquid adsorption combined with infrared spectroscopy characterization revealed that Pd@SG-ZSM-5 catalyst adsorbs much less reactant and product molecules than traditional catalyst, thereby suppressing the formation of by-products and leading to high selectivity. In chapter three, we developed a new method to encapsulate in situ produced molybdenum carbide (MoCx) in zeolite ZSM-5 for the methane dehydroaromatization (MDA) reaction. In this method, the structure-directing agent used to synthesize hierarchical zeolite ZSM-5 was utilized to reduce molybdenum precursor through a calcination process in an inert atmosphere. The zeolite subsequently underwent a secondary growth process to achieve encapsulation. The catalytic behavior of the as prepared catalyst in MDA consolidate our previous conclusion that MoCx particles outside the microporous channels can also act as the active sites for MDA, whereas it is traditionally viewed that only MoCx clusters inside the micropores are active sites. In addition, the encapsulation strategy allowed us to design experiments to answer one open question related to MDA, namely whether the Brønsted acid (BA) sites of the zeolite play a catalytic role in the conversion of methane to aromatics or only promote the dispersion of the Mo species. We encapsulated MoCx particles, which had proven to be active sites, in pure siliceous zeolite (Silicalite-1) that does not contain BA sites. The catalyst did not exhibit MDA activity even when aromatic compounds were introduced into the system by pre-adsorption or co-feeding, indicating that the BA sites are responsible for the oligomerization/cyclization step during MDA. Finally, in chapter five, we summarized the dissertation and gave our perspectives and outlooks on the further development of encapsulated catalysts based on zeolites.