Two-dimensional color Doppler tissue imaging (CDTI) has so far been used, in general, to evaluate ventricular function. This study examined if the left atrial appendage tissue velocity could reproducibly be measured with CDTI and if they have any predictive value for left atrial appendage (LAA) function and former thromboembolism. Thirty-six patients (24 women, 12 men; mean age 45 +/- 12 years; 18 AF; 11 former thromboembolic stroke) with mitral stenosis undergoing transesophageal echocardiography were examined with CDTI. Peak systolic tissue velocity (m/sec, peak systolic velocity [PSV]) was measured at the tip of the LAA in the basal short-axis view. LAA flow emptying (LAAEV) and filling (LAAFV) velocities (m/sec) were also recorded 1 cm immediately below the orifice of the appendage. Interobserver and intraobserver variabilities were determined for the PSV LAA ejection fraction was measured by Simpson's method. Mitral regurgitation, AF, transmitral mean gradient, left ventricular ejection fraction, mitral valve area, and left atrial diameter were used as a covariant for adjustment. The intraobserver and interobserver correlation coefficients for the PSV using CDTI was 0.64 and 0.60, respectively (both P = 0.01). LAAEV (0.29 +/- 0.09 vs 0.19 +/- 0.04, P = 0.001) and LAA ejection fraction (44 +/- 12 vs 29 +/- 14, P = 0.004) were found to be significantly decreased in the patients with decreased PSV (<0.05 m/sec), even after adjustment. The decreased PSV was positively correlated with the low LAAEV (<0.25 m/sec) and history of thromboembolism (r = 0.59, r = 0.38, respectively), and remained a significant determinant of the low LAAEV (OR 50.03, CI 1.46-1738. 11, P = 0.02), but not of history of thromboembolism (OR 4.29, CI 0.52-35.01, P = 0.08) after adjustment. In conclusion these results suggest that CDTI provides a reproducible method for quantification of contraction at the tip of the LAA. Decreased PSV may be predictive of poor LAA function.