Dynamic global vegetation models (DGVMs) offer explicit representations of the land surface through time and have been used to research large-scale hydrological responses to climate change. These applications are discussed and comparisons of model inputs and formulations are made among and between DGVMs and global hydrological models. It is shown that the configuration of process representations and data inputs are what makes a given DGVM unique within the family of vegetation models. The variety of available climatic forcing datasets introduces uncertainty into simulations of hydrological variables. It is proposed that satellite-derived data, validated thoroughly, could be used to improve the quality of model evaluations and augment ground-based observations, particularly where spatial and temporal gaps are present. This would aid the reduction of model uncertainties and thus potentially enhance our understanding of global hydrological change.