Abstract: Nowadays, engineering hydrology is facing many challenges such as the extrapolation or regionalization of the observations of watershed behaviors. The difficulties in producing concise, easily understood explanations of watershed behaviors have been considered as a theoretical bottleneck by the Prediction in Ungaged Basins (PUB) of the International Association of Hydrological Sciences (IAHS). An applicable solution is to develop a classification system based on dimensionless similarity indices, group watersheds into distinct groups, and then transfer observed results from gauged to ungaged basins under the frame of hydrologic similarity. In this study, it is found that hydrology does not have a theoretical system of classification and similarity as hydraulics, chemistry and biology, etc. By reviewing and comparing the similarity theory in other disciplines, we discussed the concept of hydrologic similarity, its basic methodology and three primary components, i.e., driving, structural, and hydrodynamic factors. Finally, two useful methods for hydrologic similarity (mathematical and dimensional analysis) as well as the future directions are discussed.