Estimation of liver perfusion parameters on dynamic contrast enhanced images quantitatively characterizes contrast kinetics within healthy and diseased liver as well as focal liver lesions potentially improving diagnosis and treatment monitoring. In this work we retrospectively evaluate perfusion parameters between various liver lesions and normal appearing liver by applying perfusion analysis to routine 3D dynamic contrast enhanced clinical data. 20 patients with known hepatic lesions based upon tissue sampling and follow-up imaging including metastasis, adenoma, focal nodular hyperplasia (FNH), hepatocellular carcinoma (HCC) were includeded in this study. Each patient was imaged using at 1.5 Tesla (Signa 16.0 GE) using a body phased array coil for signal reception. Dynamic spoiled gradient echo contrast enhanced imaging with a spiral acquisition lasting 60 to 80 seconds was performed using gadoxetate (Eovist), 10ml at 1ml/s followed by a 20ml saline flush. 4D images were reconstructed using PROUD [1] to obtain 288 time frames for each slice (5 mm thickness) at a temporal resolution of 0.27 sec per frame. Imaging parameters were 15º flip angle, 256 by 256 pixels, 0.8 pixel resolution per mm, repetition time of 6.0320 msec, echo time of 0.5560 msec, and imaging frequency of 63.7722 MHz. Signal intensity in each dataset was converted to relative signal enhancement and four ROIs were selected including aorta, portal vein, normal liver, and lesion. Processed images were loaded into Platform for Research in Medical Imaging (PMI 0.4) package introduced by Sourbron, et al [2]. A dual input compartment model with a single extracellular compartment was selected, since only the first 80 s of contrast enhancement were studied. [P1]How many?Table 1 provides perfusion parameters (mean, standard deviation) for normal liver vs. lesions. Extracellular volume values in the liver and metastasis are smallest followed by adenoma, HCC, and FNH. Although metastasis in this category does not exhibit a significant difference with liver, comparison of mean transit times show that metastasis is the largest followed by HCC, Adenoma, liver, and FNH. Although the difference in each category might not be significant comparing both extracellular volume and mean transit time together we see that metastasis has the smallest extracellular volume and largest mean transit time while FNH shows the opposite behavior. Comparing arterial flow fraction we can conclude that a larger fraction of plasma from portal vein is drawn into the liver (as is expected for a healthy liver) while arterial plasma flow is larger in lesions. There is a clear separation in this category between normal liver and lesions. This study assessed feasibility of quantitative characterization of normal liver vs. various hepatic lesion types. Arterial fraction and mean transit time are sensitive parameters distinguishing healthy liver from lesions although there is considerable overlap of perfusion parameters for most of the lesions. Motion artifacts due to respiratory activities during image acquisition creates high frequency oscillation on top of the enhancement curves that increases error in fitting and respective calculated tissue parameters. In addition, heterogeneity of the tissue requires careful selection of the region of interest. For instance in metastasis, signal enhancement starts at the tumor periphery. Therefore, selecting the center, ring, or combined regions will generate different results for the same case.