The Substrate Mapping and Ablation in Sinus Rhythm to Halt Ventricular Tachycardia (SMASH-VT) trial is the largest randomized trial in substrate-based ablation. We performed a retrospective analysis of patients randomized to prophylactic ablation of ventricular tachycardia to determine the predictive value of clinical and procedural variables on outcomes. In patients treated with catheter ablation, we examined predictors of ICD-therapy free survival using Cox proportional hazards models. Procedural variables tested included the scar location, number of VT morphologies (VTs) induced, tachycardia cycle length, catheter irrigation, catheter approach, procedural duration, and VT inducibility after ablation. Clinical variables including age, index arrhythmia, NYHA class, ejection fraction, prior revascularization, and baseline medication use were also analyzed. Among 64 patients randomized to ablation, 61 received the assigned therapy and complete procedural data were available for 54 patients. Thirteen percent (7 of 54) experienced ICD therapies during 2-year follow-up. Patients with subsequent ICD therapies had significantly more VTs induced during the ablation procedure than those without (3.9 +/- 2.1 vs 1.9 +/- 1.8, P = 0.05). The hazard ratio for each additional VT induced was 1.51 (95% CI 1.07-2.13, P = 0.02). Two-year Kaplan-Meier event-free survival rates were 96% for 0-1 VTs induced, and 78% for two or more. The use of irrigated catheters was not predictive of ablation success. In this small retrospective analysis, the number of VTs induced during the procedure was predictive of 2-year outcomes. This likely reflects a more complex arrhythmia substrate in patients who fail ablation.