Non-contact (NC) mapping efficiently acquires global activation wavefronts and is potentially more accurate than contact (C) mapping due to elimination of C-related artifact. Common areas of conduction velocity (CV) deceleration during sinus and paced rhythms have been used to characterize the atrial substrate, however, such analysis with C mapping is time-consuming. Compare CV using C local activation timing (LAT) and NC LAT maps. Retrospective analysis of CV of NC LAT maps (SuperMap, Acutus Medical) obtained from 22 clinical pts were compared to CV of C LAT maps created from unique C EGMs concurrently acquired with the AcQMap Catheter. NC LAT maps were created from charge density EGMs inversely computed from NC measurements. For both C and NC datasets, CVs were computed using a polynomial surface fit. C and NC CV vectors were compared for both direction and magnitude using Bland-Altman analysis. C and NC derived CV comparisons (n=780) were made from 29 maps of SR, AT/AFL or pacing. Mean acquisition time/map was 2 min 40s ± 1 min. Figure 1 shows a C and NC generated LAT map with corresponding CV overlaid and Bland-Altman agreement limits. Bland-Altman analysis showed bias between both methods was small with the confidence interval for C CV magnitude at 6.0 ± 17.0 cm/s and for C CV direction being 9.8±16.6°. NC LAT maps were efficiently acquired, and NC CV measurements were consistent when compared to C CV. Results of this study are consistent with analysis of CV from a high-density grid compared to optical mapping (Masse et al, 2016). Accuracy and time efficiencies achieved by NC mapping may facilitate more widespread use of substrate characterization.