The carbon cycle in forest ecosystems is an important part of the global terrestrial ecosystem carbon cycle, and models have become a necessary means used to study the forest carbon cycle. Forest carbon cycle models can be classified into statistical models and process models, and the latter have gradually become dominant due to their complete theoretical framework, rigorous structural analysis, and clear process mechanism. This paper reviews the studies on forest carbon cycle process models on a large scale from three aspects:geochemical process models, terrestrial physical process models, and biological process models. Their main features, advantages, and application status are summarized, and the limitations of these models are also discussed systematically. Geochemical process models focus on the cycle process of important elements (such as carbon, nitrogen, and water) among the vegetation, litter, and soil organic matter, which can be used to simulate the forest carbon balance, vegetation productivity, and nutrient utilization. Terrestrial physical process models focus on the process of energy and momentum exchange between vegetation and the external environment under different atmospheric conditions,which emphasize the interaction between climate and vegetation. Meanwhile, complex processes are considered by terrestrial physical process models, including reflection, absorption, scattering, transmission, and other processes. Biological process models, focus on the analysis of vegetation composition, distribution characteristics, and dynamic changes under different environmental conditions. Biological process models can be further divided into biogeographic models, light-use efficiency models, and dynamic vegetation models. In this review, the existing problems associated with the simulation of forest carbon cycle processes are discussed, such as the uncertainty of input data and model mechanism, scale effect, and topographic effect, which need to be further studied. One of the main problems is the uncertainty of input data, which affects the accuracy of forest carbon cycle process simulation. Furthermore, the uncertainty of model mechanism also makes the simulation difficult. Another issue is the scale effect associated with the mismatch among the observation, model, and surface process scales. The scale effect may cause contrasting results in the simulation of forest carbon cycle processes. Note that there exists obvious topographic effects on simulation models in complex terrains. Finally, the prospects of implementing these models into future studies are discussed. When ignoring the feedback effect of vegetation on atmosphere, geochemical process models are suitable for the simulation of carbon budget on a regional scale, and the biogeographic models are suitable for the study of vegetation distribution characteristics on a global scale. However, under the current background of global change, in order to meet the various needs of forest carbon cycle simulation, such as carbon budget, evapotranspiration process, vegetation succession process, and vegetation feedback to the atmosphere, the terrestrial physical process models and the dynamic global vegetation models will be used as the mainstream research into forest carbon cycle process models in the future. This review can be served as a reference when selecting models for forest ecosystem carbon cycle simulation at various spatial scales.