Molecular recognition is one of the fundamental events in biological systems, as typified by enzymes that enable highly efficient and selective catalytic reactions through precise recognition of substrate(s) and cofactor(s) in the binding pockets. Chemists therefore have long been inspired by such excellent molecular systems to develop various synthetic receptors with well-defined binding sites. Their effort is currently being devoted to the construction of not only molecular receptors but also self-assembled host compounds possessing connected cavities (pores) in the crystalline frameworks to rationally design functional porous materials capable of efficiently adsorbing molecules or ions at binding sites on the pore walls. However, it is still challenging to design multiple distinct binding sites that are precisely arranged in an identical framework, which is currently one of the most important targets in this field to realize elaborate molecular systems beyond natural enzymes.In this Account, we provide an overview of porous crystals with well-defined molecular recognition sites. We first show several strategies for arranging macrocyclic binding sites in crystalline frameworks such as metal-organic frameworks, porous molecular crystals, and covalent organic frameworks. Porous metal-macrocycle frameworks (MMFs) that we have recently developed are then described as a new type of porous crystals with well-defined multiple distinct binding sites. The MMF-1 crystal, which was developed first and is composed of four stereoisomers of helical PdII3-macrocycle complexes, has one-dimensional channels with dimensions of 1.4 nm × 1.9 nm equipped with enantiomeric pairs of five distinct binding sites. This structural feature of MMF-1 therefore allows for site-selective and asymmetric arrangement of not only single but also multiple guest molecules in the crystalline channels based on molecular recognition between the guests and the multiple binding sites. This characteristic was also exploited to develop a heterogeneous catalyst by non-covalently immobilizing an organic acid on the pore surface of MMF-1 to conduct size-specific catalytic reactions. In addition, adsorption of a photoreactive substrate in MMF was found to switch the photoreaction pathway to cause another reaction with the aid of photoactivated PdII centers arranged on the pore walls. Furthermore, the dynamic, transient process of molecular arrangement incorporated in MMF-1 has been successfully visualized by single-crystal X-ray diffraction analysis. The formation of homochiral MMF-2 composed of only (P)- or (M)-helical PdII3-macrocycle complexes is also described. Thus, macrocycle-based porous crystals with a complex structure such as MMFs are expected to serve as novel porous materials that have great potential to mimic or surpass enzymes by utilizing well-defined multiple binding sites capable of spatially arranging a catalyst, substrate, and effector for highly selective and allosterically tunable catalytic reactions, which can be also visualized by crystallographic analysis because of their crystalline nature.