Ventricular tachycardia (VT) is a life-threatening arrhythmia, which can be treated by catheter intervention. Accurate identification of the underlying reentrant circuit is often challenging, yet it is key to successful ablation of the VT. In practice, the cardiologist often uses electrocardiography (ECG) data provided by various catheter mapping techniques, including parameters acquired during sinus rhythm (voltage maps, presence of fragmented/late potentials) and during controlled pacing from different sites of the ventricle, so-called pace-mapping. A novel method is presented here to automatically extract the key information from pace-mapping data with automated detection of paced heartbeats from the surface ECG signals, using wavelet detection of pacing spikes and combined time/energy criteria, and automated delineation of paced beats, QRS peak, and QRS onset. This allows the generation of correlation gradient maps (indicating QRS morphology changes as the catheter is moved) and stimulus-to-QRS maps (sQRS, indicating the delay between pacing and activation of the healthy myocardium). The delineator is shown to be in good agreement with manual annotations from experts in a retrospective study of 18 VT ablation procedures. Paced-QRS detection had 95.2% sensitivity and 98.4% positive predictive value. Resulting sQRS maps had a mean absolute error of 11.1 ms, which was in the same range as the inter-observer errors (9.7 ms). The automatic processing drastically reduces the need for manual annotations. Therefore it makes it feasible to process and visualize, during the procedure, all the relevant parametric maps, which can be analyzed jointly to identify VT circuits and corresponding ablation targets.