BACKGROUND: Quantification of left ventricular (LV) dyssynchrony allows for objective measures of resynchronization therapy (CRT) effectiveness. We tested the hypothesis that site of LV pacing, fusion beats and baseline contractility alter contraction synchrony as quantified by regional and global measures of LV performance. METHODS AND RESULTS: In 8 open-chested pentobarbital-anesthetized canine preparations we compared the effects of right atrial (RA), RA-high right ventricular (RV) free wall, as a model of left bundle branch block contraction pattern, RA-LV apex (LVa), RA-LV free wall (LVfw), and RA-RV-apical LV (CRTa) and RA-RV-free wall LV (CRTfw), as CRT. LV pressure-volume loops recorded using high-fidelity pressure and conductance catheters and echocardiographic angle-corrected color-coded strain imaging of mid-LV short axis views analyzed radial strain from six segments. To control for contractile state esmolol-induced beta blockage was studied, and in 5 dogs to control for RA and ventricular pacing fusion beat artifacts, repeat studies were done following AV node ablation. RA-RV pacing reduced stroke work (SW) (57±18 to 33±13* mmHg·mL,*p<0.05 vs RA pacing), decreased LV end-diastolic volume and induced marked radial dyssynchrony (maximal time difference between peak segmental strain) from 31±15 to 234±60* ms. Changes in radial dyssynchrony correlated significantly with changes in SW (r=-0.53, p<0.01). Dyssynchrony improved with both CRTa and CRTfw (69*±31 and 98*±63 ms, respectively) while SW only improved with CRTa (62±22* and 37±13 mmHg·mL, respectively * p<0.05 vs RV pacing). CRTa also tended to increased LV end-diastolic volume over RA-RV. Esmolol slowed HR from 118±10 to 108±10 beats/min* and tended to decrease contractility (end-systolic elastance (Ees) from 12.1±7.9 to 8.9±3.9 mmHg/ml, p=0.167) but did not alter the degree of RV-pacing induced dyssynchrony. AV ablation had no effect on the observed apical and free wall contraction differences seen during baseline conditions. CONCLUSION: Although both CRTa and CRTfw reduced contraction dyssynchrony, CRTa tended to improve global LV performance more by increasing end-diastolic volume. Thus, CRT may improve global LV performance differently, depending on the LV pacing site.