To relieve and solve the energy problem, effective methods to use solar energy must be built up. To be more specifically, we need figure out how to utilize sunlight for water splitting reaction, giving rise to hydrogen as clean energy, and photoreduction of carbon dioxide, leading to the formation of useful liquid products (e.g., HCOO−, HCHO and CH3OH) or gaseous products (e.g., CH4 and CO). As a class of distinguished and unique materials, metal-organic frameworks (MOFs) have drawn a lot of attention, considering that they display special physical and chemical properties such as exceedingly high surface areas, designable and controllable cavities, different mechanisms of photo-induced electrons transfer, and moreover photoactive parts can be easily introduced into MOFs by either encapsulating dye molecules into the cavities or constructing the frameworks with optically active bridging ligands or metal nodes. In this review, we have commented on the challenges in this field, summarized the unique advantages and inherent merits of MOFs as the emerging materials, and pointed out the opportunities and development strategies of MOFs for their applications in photocatalysis. Firstly, we have introduced MOFs′ concepts and features, distinguishing them from other porous materials, and their advantages in photocatalysis. MOFs are crystalline porous materials formed from ligands (including metalloligands) and transition-metal nodes. The structures of MOFs are of facile design, and can be further modified through post-synthetic methods. Some MOFs display high thermal and chemical stability, which can be stable up to 500°C and resist a variety of reaction media either organic solvents or aqueous, even in acidic and basic solutions. MOFs can be photoresponsive through light absorption by the organic linker, the metal oxide nodes or the photoactive species entrapped in the voids. Photoexcitation of the light absorbing units in MOFs generates the excited state, which might induce photocatalytic activity. Next, we have classified photocatalytic MOFs into three types, including (1) metal-oxo clusters as semiconductor dots, (2) ligands/metalloligands as photocatalysts, and (3) photocatalytic species (nano- particles, polyoxometalates, nano-composites, and etc.) encapsulated into the pore, and discussed their applications in photocatalysis in details. As for type I, metal-oxo clusters, especially Zr-O or Ti-O clusters, as the nodes have been assembled into MOFs. Upon the absorption of photons with the energy greater than the bandgap of the ligand, a ligand-to-metal charge-separation state was generated, resulting in photocatalytic activity. In type II, a few molecular photocatalysts based on metal-polypyridine complexes, usually being Ru and Ir complexes, metalloporphyrins and organic dyes have been incorporated into MOFs to afford photocatalysts under visible light. Considering type III, photoactive species, including polyoxometalates and metal nanoparticles (e.g., Pt, Pd, Au, and Ag NPs), have been doped into the cavities of MOFs. In addition, integration of an inorganic semiconductor with a MOF gives rise to a composite photocatalyst, which combines the advantages of both materials and then results in higher efficiency, selectivity and stability (especially low metal leaching and recyclability). Finally, we have provided our perspectives for the future of MOFs as photocatalysts. MOFs have displayed the potentials in photocatalysis, but there still exist large improvement spaces. The relatively low stability of MOFs compared to inorganic semiconductors limits their applications for practice. In most reports of MOF photocatalysts, sacrificial agents are required, which isn′t consistent with the sustainable development concept. MOFs with strong absorption of visible light, long lifetime of excited state, high product selectivity and stability are in pursuing.