In recent years, the concept of ecosystem-based fisheries management (EBFM) has been widely recognized. The achievement of EBFM goals requires a deep understanding of the structure, function, and mechanisms of the ecosystem, as well as appropriate modeling methods to evaluate and predict the changes in biological resources and ecosystems. Fishery ecosystem dynamic models have developed in this context, aiming to investigate the structure and function of fishery ecosystems, and to evaluate the effects of environmental changes and human intervention. This category of models has gradually become the major tool for information synthesis, integration, and prediction in ecosystem research. The fishery ecosystem dynamic models have developed rapidly in the past 20 years, whereas due to their wide diversity, there are substantial difficulties in the selection, construction, and application of those models. This study traces the development of fishery ecosystem dynamic models, and briefly introduces representative models at the population, community, and ecosystem levels, respectively. The fishery ecosystem dynamic models bridge the gap between microcosmic processes and macroscopic phenomena, and serve as a crucial tool for revealing the regulatory mechanisms of ecosystems. In the future, the fishery ecosystem dynamic models need further development, and should be adjusted and improved based on our understanding of the ecosystem, technical foundations, and data sources, in order to provide supports for the practice of EBFM.