We present a method that leverages 3D image-based navigators (iNAVs) for nonrigid motion correction in free-breathing, non-contrast-enhanced renal angiography scans. We begin by performing an ROI-based analysis of 3D iNAV motion, with ROI selection based on published biomechanical simulations of the renal arteries during respiration. Then, we demonstrate that localized motion estimates derived from different ROIs agree with the findings of the simulations. Finally, we combine the extracted motion information with an autofocusing technique for respiratory motion compensation. Across all patient studies, the proposed method significantly improves the depiction of the renal arteries as compared to 3D translational motion correction.