Agricultural multi-pond systems, typical of the landscape of China, have been documented historically for 3,000 years. As small-water conservancy projects, multi-pond systems are widely distributed in southern China. They comprise a large artificial irrigation network system, composed of many tiny ponds, scattered in agricultural fields and connected by ditches and streams. They perform a key function in hydrological regulation and mass cycling. In recent decades, there have been many studies on the ecological process of multi-pond systems. Based on the review of multi-pond systems and their hydrological processes, the main study of multi-pond systems can be divided into two aspects: ecosystem and landscape scales. (1) For ecosystem scales, the change of hydrological processes and the nutrient recirculation were developed. Local hydrological cycling was changed, and the ponds could control flooding rate and increase terminal dwell times of runoff. However, the multi-pond system has other ecological functions that clean up water quality, i.e., sediment and nutrient retention. (2) For landscape scales, we reviewed examples of empirical models and potential mechanistic modeling tools that could be applied to further advance scientific understanding. Statistical models allow a steady state analysis that links landscape characteristics to watershed hydrology and water quality data. However, mechanistic watershed models are structured to dynamically link landscape features to downstream hydrological and biogeochemical processes. Here we discuss a subset of watershed models readily adaptable to address multi-pond connectivity questions, including the Soil and Water Assessment Tool (SWAT), Hydrological Simulation Program FORTRAN (HSPF) model, DRAINmod for WATershed model (DRAINWAT), and TOPMODEL model. In watersheds with a sufficiently dense distribution of ponds, the hydrologic functions they provide could have important implications for flood regulation and mitigation of the future effects of climate and land use change. With the recognized function of multi-pond systems, the focus was on the hydrological module of ponds, and this was developed to enhance the models above. Based on the review of multi-pond systems, there were some shortages of studies between multi-ponds and their ecological process, including the following aspects. Firstly, analyses on ecological scales were based on the cycling mass in one pond and typically used field measurements to elucidate the source of a dissolved element in surface waters by accounting for all known sources, losses, or sinks. However, pond connectivity was often overlooked. In a landscape context, thousands of ponds are connected in a large network to provide ecological function at a regional scale. Secondly, several empirical methodologies hold the potential to improve the accuracy of mechanistic modeling approaches by providing parameter value estimates. Empirical methodologies, such as similar statistical approaches, could therefore highlight important factors associated with ponds (e.g., maximum and minimum volumes and surface depression storages, distance to stream network) that influence flow and highlight important variables to accurately parameterize the distributed landscape-scale mechanistic models. Thirdly, the research of multi-pond systems in China always depended on the foreign models, which cannot accurately reflect the characteristics of systems in China.