Optimization of the interventricular pacing delay (VV) in cardiac resynchronization therapy is time-consuming and not routinely performed. The aim of the present study was to compare the acute hemodynamic response obtained by different VV programming methods. Several methods for optimizing the VV using electrocardiographic or echocardiographic measurements were performed. The effect of programming an empirical prefixed VV of 0 ms was also evaluated. Invasive first derivative of left ventricular (LV) pressure over time (dP/dt max) was measured at several VV values, and the hemodynamic response that could be obtained by each noninvasive VV selection method was extrapolated from the curve of LV dP/dt max versus VV. The study included 25 patients (80% men, age 66 ± 9 years, 44% ischemic). The maximum achievable LV dP/dt during biventricular pacing was obtained by a median left ventricular preactivation of 30 ms and increased the baseline unpaced LV dP/dt from 774 ± 181 to 934 ± 179 mm Hg/s (p <0.001). The noninvasive optimization method selected the VV leading to the narrowest QRS measured from the earliest deflection and obtained the smallest difference with regard to the maximum achievable LV dP/dt. Furthermore, of all the VV optimization methods tested, this was the only 1 that significantly improved on the hemodynamic response obtained by programming a predefined VV of 0 ms in all patients (925 ± 178 vs 906 ± 183 mm Hg/s; p = 0.003). In conclusion, achieving the narrowest QRS measured from the earliest deflection obtained a better acute hemodynamic response than the other VV optimization methods. It also improved the response obtained by default simultaneous biventricular pacing, although this improvement was limited in magnitude.