Arterial diameter change is related to distending blood pressure and is used in estimation of arterial stiffness parameters. A common technique to track the arterial walls is by integration of wall velocities estimated by different methods using cross correlation or tissue Doppler. However, because of the high pulse repetition frequency and the need for separate pulsing sequences, the B-mode image quality is affected. We have previously developed a fast algorithm for direct measurement of lumen diameter using B-mode images. In this study we have improved the technique to be more robust and also implemented measurements of diameter change, maximum differential wall velocity and relative diameter change of the common carotid artery noninvasively in vivo. The influence of the lateral width of the region of interest (ROI; 1 pixel, 0.1 mm, 0.5 mm, 1 mm, 2.5 mm and 5 mm) was evaluated. Using the optimal lateral width (2.5 mm), the systematic and random differences between two consecutive measurements were 21 microm and 105 microm, respectively, for lumen diameter measurement; -7 microm and 39 microm, respectively, for lumen diameter change measurements; -0.2 mm/s and 0.9 mm/s, respectively, for maximum relative wall velocity; and -0.2 % and 0.8 %, respectively, for measurements of the relative diameter change. The coefficient of variation (CV) was 1.9%, 5.2%, 7.9% and 6.0%, respectively. The study indicates that the reproducibility is sufficient for in vivo studies when the width of the ROI is 1.0 mm or wider.