In order to evaluate CrIS geolocation at a sub-pixel level, a method is developed by using the collocated radiance measurements from Visible Infrared Imaging Radiometer Suite (VIIRS) infrared bands. Specifically, an algorithm has been developed to accurately determine the footprint shapes of CrIS FOVs. The VIIRS pixels are then spatially averaged to compare with CrIS measurements, which are convolved with VIIRS spectral response function to generate VIIRS band radiances. By shifting the VIIRS pixel toward along- and cross-track direction, the perfect collocation position can be indentified to quantify the geolocation bias of CrIS geolocation FOVs. The poster overviews the method, summarizes the first year of geolocation analysis, and provides the assessment results on CrIS geolocation. More important, the proposed approach allows an operational, long-term characterization of the CrIS geolocation and enables individual CrIS observations to be geolocated to the sub-pixel accuracy that is required for CrIS data application.