Speckle tracking echocardiography (STE) has the potential to detect the heterogeneous initiation of active myocardial contraction, which is the primary cause of left ventricular (LV) systolic dysfunction in patients with mechanical dyssynchrony. This study was designed to assess the ability to predict response to cardiac resynchronization therapy (CRT) of STE-derived dyssynchrony parameters in comparison with invasive hemodynamic assessments. Thirty patients referred for CRT were studied. The time difference of first peak (Td-(first peak)) and the maximum peak (Td-(max peak)) on the radial strain-time curves of the earliest and the latest segments were measured as the dyssynchrony parameter. Peak positive dP/dt (dP/dt(max)) was measured as the indicator of global LV systolic performance. CRT responders were defined as patients with LV end-systolic volume reduction greater than 15% at 3 months after CRT. CRT increased the dP/dt(max) compared with the baseline study (P < .001). Percent changes in the dP/dt(max) (dP/dt(max)) were significantly correlated with Td-(first peak) (R = 0.743, P < .001), but weakly correlated with Td-(max peak) (R = 0.390, P = .03). Twenty patients (66%) were identified as chronic CRT responders. Receiver operating characteristics analysis revealed that Td-(first peak) shared a similar ability with dP/dt(max) to detect chronic responders (Td-(first peak) >167.0 ms, area under the curve [AUC] 0.945; dP/dt(max) >16.2%, AUC 0.934) compared with Td-(max peak) (>194.5 ms, AUC 0.820). STE-derived Td-(first peak) showed a reliable ability to predict the acute and chronic response to CRT as well as dP/dt(max).