Dual-source computed tomography (CT) can evaluate left ventricular (LV) dyssynchrony, myocardial scar, and coronary venous anatomy in patients undergoing cardiac resynchronization therapy (CRT). We aimed to determine whether dual-source CT predicts clinical CRT outcomes and reduces intraprocedural time. In this prospective study, 54 patients scheduled for CRT (mean age 63 ± 11 years; 74% men) underwent preprocedural CT to assess their venous anatomy as well as CT-derived dyssynchrony metrics and myocardial scar. Based on 1:1 randomization, the implanting physician had preimplant knowledge of the venous anatomy in half the patients. In blinded analyses, we measured time to maximal wall thickness and inward wall motion to determine (1) CT global and segmental dyssynchrony and (2) concordance of lead location to regional LV mechanical contraction. End points were 6-month CRT response measured using heart failure clinical composite score and 2-year major adverse cardiac events (MACE). There were 72% CRT responders and 17% with MACE. Two wall motion dyssynchrony indices-global wall motion and opposing anteroseptal-inferolateral wall motion-predicted MACE (P < .01). Lead location concordant to regions of maximal wall thickness was associated with less MACE (P < .01). No CT dyssynchrony metrics predicted 6-month CRT response (P = NS for all). Myocardial scar (43%), posterolateral wall scar (28%), and total scar burden did not predict outcomes (P = NS for all). Preknowledge of coronary venous anatomy by CT did not reduce implant or fluoroscopy time (P = NS for both). Two CT dyssynchrony metrics predicted 2-year MACE, and LV lead location concordant to regions of maximal wall thickness was associated with less MACE. Other CT factors had little utility in CRT.