There is a critical need for a noninvasive method to determine the optimal treatment for patients with deep vein thrombosis (DVT), the 3rd most common cardiovascular disease. We have implemented multiparametric MRI to evaluate spatial and temporal changes in thrombus composition using two murine models of DVT. We show that T2- and T2*-weighted MRI detects blood-flow dependent variations in thrombus composition. Classification in fully occlusive thrombi indicated an increased inflammatory response and more rapid thrombus organization when compared to thrombi developed in the presence of blood flow. This work provides foundational methodology that could eventually inform optimal DVT treatment planning.