Paradoxically, instead of reducing perfusion, these therapies can result in the “normalization” of vessels and improved perfusion. This is important as it can increase the efficacy of subsequent radiation or drug delivery, resulting in better response. Thus, combination therapies with an anti-angiogenic agent followed by a targeted or cytotoxic agent are becoming an important anti-cancer paradigm. DSC-MRI is a rapid imaging technique that can yield quantitative estimates of microvasculature changes, which may aid in identifying the normalization time window for preclinical chemotherapy studies. This relies on compartmentalization of the CHIR-99021 contrast agent while a susceptibility difference can be induced between intravascular and extravascular spaces. Signal loss caused by spin dephasing during the first pass of CA circulation is associated with the magnetic field distortions in the vicinity of vessels. The susceptibility differences could also be affected by vascular permeability and tortuosity. Although a vast number of studies have used this technique, the issue of reproducibility has often been underappreciated Only a few studies have addressed the reproducibility of DCE-MRI. Reproducibility is especially important for the application of assessment of tumor response to treatment. Additionally, there are few studies comparing the reproducibility of different MW contrast agents. This study demonstrated the reproducibility of DCE-MRI with CAs that have small and large molecular weights, and using standard MRI acquisition and analysis protocols in a breast xenograft tumor model. A major finding in this work was that the VIF was highly variable and needs to be calculated for each experimental setup. An automated VIF ROI selection method was developed that averaged time activity from voxels that were identified using pre-defined kinetic thresholds. The detection of a reproducible VIF is crucial to obtain reproducible estimates of kinetic parameters. Multiple factors for VIF estimation that may induce errors include partial volume effects, low temporal and spatial resolution, low contrast to noise ratio and in many DCE-MRI cases, a poor choice of major arteries in the field of view and motion artifacts. In our study, the VIF was estimated using the contrast agent enhancement data from three T1-weighted images of the thigh. Multiple ROIs were selected using an automatic detection algorithm that traces the signal changes of each pixel through all slices. Potential problems with motion artifacts and low SNR were mitigated by using a temporal spline-fit function and a spatial Gaussian kernel to smooth the data for pixel-wise analyses. We had hypothesized that the larger CA would be more reproducible, as it is more sensitive to permeability and less sensitive to flow effects.