Because it appears to be less affected by changes in preload, mitral annulus Doppler tissue imaging (DTI) has been proposed as an alternate mean of identifying pseudonormal patterns of left ventricular filling. We thus studied the practical implications of DTI in 40 patients classified according to the Canadian Consensus on Diastolic Function (9 control participants, 9 with impaired relaxation, and 22 pseudonormal participants). Using DTI, the early diastolic velocity (Ea) was the most reproducible parameter whereas the late diastolic velocity (Aa) and Ea/Aa ratio varied significantly. Nonetheless, Ea missed 23% of pseudonormal participants and its sensitivity, specificity, and positive and negative predictive values to identify diastolic dysfunction were 81%, 89%, 96%, and 57%, respectively; improving to 94%, 89%, 97%, and 80%, respectively, if used during Valsalva's maneuver. Thus, DTI is not totally preload independent and should be interpreted in light of the other Doppler parameters and the use of Valsalva's maneuver. Moreover, consistent with recent studies, these results suggest that the threshold value of Ea used to identify diastolic dysfunction should be approximately 12.5 cm/s.