Risk for ventricular arrhythmias (VA) following cardiac resynchronization therapy (CRT) has been associated with ischemic disease/scar, sex, and possibly left ventricular mass (LVM). To evaluate sex differences and baseline/post-implant change [Δ] of LVM on VA risk after CRT implant among patients with non-ischemic cardiomyopathy (NICM) and left bundle branch block. Among patients meeting the criteria, baseline and follow-up echocardiographic images were obtained for LVM assessment. VA events were reported from device diagnostics and therapies. VA risk was stratified by ROC (Youden-index cut-point) for baseline LVM and ΔLVM, and baseline patient characteristics using a multivariable Cox regression model. 118 patients (71[60.2%] female, age 60.5 ±11.3 years, LVEF 19.2 ±7.0%, QRS 165.6 ±20 ms, LVM 313.9 ±108.8 g) were enrolled and followed for median 90 (IQR 44-158) months. Thirty-five (29.6%) patients received appropriate shocks or anti-tachycardia pacing at a median of 73.5 (IQR 25-130) months post-implant. Males had a higher VA incidence (male 18/47 [38.3%] vs. female 17/71 [23.9%], P=0.02). Baseline LVM >308.9g separated patients with higher VA risk (P=0.001). Less than a 20% decrease in LVM increased VA risk (P<0.001). Baseline LVM was the only baseline characteristic predicting VA events in the Cox regression model (Hazard ratio 1.01 [95%CI, 1.001-1.009], Log-rank P=0.003). Sex differences in VA risk were eliminated by the baseline LVM parameters. VA risk after CRT in NICM was associated with baseline LV >308.9g and a decrease in LVM ≤ 20%, without sex differences.