In DDD-pacemaker patients with high degree AV block, Doppler echocardiography of transmitral blood flow can be used to find the individually optimal AV delay (AVO) for left heart AV synchronization. This study tried to validate a Doppler method (ECHO) recently proposed to optimize left ventricular filling by comparing it to stroke volume data derived from impedance cardiography (ICG). It should be further elucidated if optimizing the AV delay (AVD) by means of this method is superior to fixed AVD settings and which differential AVD (pace-sense-offset) should be programmed for atrially triggered (ATP) and AV sequential (AVP) pacing, respectively. AVO as measured in 53 patients showed a linear correlation between ECHO and ICG for both ATP (r = 0.66, P < 0.00001) and AVP (r = 0.53; P < 0.005). The mean deviation in AVO between ECHO and ICG was +/- 26 ms (ATP) and +/- 30 ms (AVP), respectively, with a tendency to longer AVDs with the Doppler method. ECHO limitations could mainly be attributed to: (1) restrictions of AVD programming options (which may be compensated for by slight modification of the proposal); and (2) to pathophysiological mechanisms that alter mitral valve dynamics. Optimization of the AVD by Doppler produced a stroke volume that was significantly higher (19%) than with a fixed AVD (150 ms in ATP; 200 ms in AVP). There was a wide scatter in pace-sense-offsets between-7 and 134 ms, which was reflected by both methods. It is concluded that AVO determinations by ECHO are valid provided that methodological pitfalls and limitations caused by the disease are recognized. Tailoring AVD with respect to diastolic filling improves systolic function and is superior to nominal AVD settings. Fixed differential AVDs as offered by some manufacturers are far from being physiological. Thus modern pulse generators should offer free programmability over a wide range of AV delays.