Recently, there has been great interest in the use of ultrasound contrast to generate quantitative information about tissue perfusion. Using low MI pulse sequences and a clearance — refill approach, it is possible to create quantitative time-to-refill maps of tissue correlating to blood perfusion rate. One limitation of 2D perfusion imaging is that only information from a single slice of tissue is gathered. Vascular inhomogeneity, common in many tumor types, makes perfusion estimates inconsistent unless the same region is imaged repeatedly. Our objective was to evaluate errors in 2D quantitative in-vivo perfusion estimates due to differences in transducer positioning compared to 3D acquisitions. We also examined the effect of contrast dose on perfusion estimates. Destruction-reperfusion imaging was performed with parametric mapping of refill times and image alignment to correct for tissue motion. Images were acquired in rat kidneys using a Siemens Sequoia 512. 3D images were generated by mounting the transducer to a computer controlled linear motion stage. Changes in perfusion times were examined as a function of contrast dose, imaging plane, and in response to the vasoactive drug dopamine. Our results showed that we could differentiate perfusion changes before and after administration of dopamine, and that 3D images were more consistent than 2D acquisitions. We also observed bias caused by contrast agent infusion rate using our time-to-refill algorithm; however, in the range of 2.7 × 108 and 3.9 × 108 bubbles/min the results obtained were consistent to within 3%. 3D perfusion imaging demonstrated a significant reduction in error caused by transducer positioning and improves the reliability of quantitative perfusion time estimates in the rat kidney model.