As metabolomics datasets are becoming larger and more complex, there is an increasing need for model-based data integration and analysis to optimally leverage these data. Dynamic models of metabolism allow for the integration of heterogeneous data and the analysis of dynamic phenotypes. Here, we review recent efforts in using dynamic metabolic models for data integration, focusing on approaches based on ordinary differential equations that are applicable to both time-resolved and steady-state measurements and that do not require flux distributions as inputs. Furthermore, we discuss recent advances and current challenges. We conclude that much progress has been made in various areas, such as the development of scalable simulation tools, and although challenges remain, dynamic modeling is a powerful tool for metabolomics data analysis that is not yet living up to its full potential.