Non-invasive characterization of complex atrial tachyarrhythmias, regarding atrial flutter (AFL) and focal atrial tachycardia (AT) is challenging. Conventional electrocardiogram (ECG) faces several limitations. Planning a preprocedural strategy before AFL-AT catheter ablation (CA) is needed. To evaluate the diagnostic capacity of a novel non-invasive electrocardiographic imaging (ECGI) system that does not require previous CT or MRI cardiac imaging. 42 patients (27 males, CHA2DS2-VASc score 2±1, LVEF 54.5 ± 11.0 %, and LA area 31.3 ± 7.0 cm2) undergoing CA for AT-AFL were prospectively included in the study. Before the procedure, the ECGI system estimated the patient atria geometry using a 3D torso reconstruction and an artificial intelligence-based algorithm. Multiple electrocardiograms were recorded using a 64-electrodes vest during the CA procedure. The different diagnostic capacities of the periprocedural ECG and ECGI were assessed compared with endocavitary electroanatomical mapping (EAM) regarding three endpoints: (1) the involved atrial cavity, (2) the mechanism (focal or macro-reentry), and (3) the ablation target site defined as the area where the earliest activation was located for focal arrhythmias, and the precise anatomical pathway for macro re-entrant tachyarrhythmias. Regarding the ECG, endpoints were assessed based on validated algorithms by 2 different observers blinded from the diagnosis. 48 atrial tachyarrhythmias were evaluated and its characterization using this novel ECGI system was possible in 94% of the cases: directly for 36 arrhythmias and after decreasing ventricular response using vagal maneuvers for 9. ECGI obtained global accuracy to identify the involved atrial cavity of 92%, the mechanism of 90%, and the ablation target of 83% (Table). In challenging cases, such as in patients with a history of CA for atrial fibrillation and in patients without a final diagnosis of typical counterclockwise AFL, the accuracy of ECGI was consistent compared to the whole study population (Table). The ECGI diagnostic capacity significantly outclassed the ECG regarding the identification of the involved cavity (p<0.001), the mechanism (p<0.05), and the ablation target (p<0.001). ECGI could have prevented all unnecessary transseptal punctures (n=7), representing 28% of the overall punctures. This novel non-invasive ECGI system accurately characterizes complex atrial tachyarrhythmias (Figure) without needing previous CT or MRI cardiac imaging.