The sampling signals from different electronic transformers need to be synchronized in the merging unit. However, when harmonics, especially when the higher harmonics, exist in the signal, it’s difficult for the existing methods to restore the original data at the sampling interval accurately, leading to a large error during the synchronization process. To solve this problem, an improved sampling data synchronization method is studied. Based on the cubic spline interpolation theory, the global feature of sampling date is analyzed, and a continuous piecewise cubic function is used to describe the waveforms between the sampling points to improve the re-sampling accuracy. Then, based on the principle of interpolation algorithm and the sampling characteristics, the normalization process, matrix transformation, convergence analysis and other methods are used to reduce the amount of calculations. Finally, through simulation and verification, this method can ensure that the resampling data no higher than the 21th harmonic is not distorted, and the calculation time is only about 10% of the equal precision interpolation algorithm. In summary, this method can not only improve the accuracy of sampling data synchronization with the influence of higher harmonics, but also solve the limitation of high precision interpolation operation. Therefore, it is valuable for engineering applications. This work is supported by National Key Research and Development Program of China (No. 2017YFB0903100), National Natural Science Foundation of China (No. 5177145), and Science and Technology Program of State Grid Corporation of China (No. 521104170043).