Atherosclerotic change of the arterial wall leads to a significant change in its elasticity. For assessment of elasticity, measurement of arterial wall deformation is required. For motion estimation, correlation techniques are widely used, and we developed a phase-sensitive correlation method, namely, the phased-tracking method, to measure the regional strain of the arterial wall due to the heartbeat. Although phase-sensitive methods using demodulated complex signals require less computation in comparison with methods using the correlation between RF signals or iterative methods, the displacement estimated by such phase-sensitive methods are biased when the center frequency of the RF echo apparently varies. One of the reasons for the apparent change in the center frequency would be the interference of echoes from scatterers within the wall. In the present study, a method was introduced to reduce the influence of variation in the center frequencies of RF echoes on the estimation of the artery-wall strain when using the phase-sensitive correlation technique. The improvement in the strain estimation by the proposed method was validated using a phantom. The error from the theoretical strain profile and the standard deviation in strain estimated by the proposed method were 12.0% and 14.1%, respectively, significantly smaller than those (23.7% and 46.2%) obtained by the conventional phase-sensitive correlation method. Furthermore, in the preliminary in vitro experimental results, the strain distribution of the arterial wall well corresponded with pathology, i.e., the region with calcified tissue showed very small strain, and the region almost homogeneously composed of smooth muscle and collagen showed relatively larger strain and clear strain decay with respect to the radial distance from the lumen.