Abstract: Upward continuation is one of the important methods used to separate gravity anomalies. However, how to quantitatively select an appropriate upward-continuation height has always been a problem in the application of this method. Given this, this paper proposes a curvature analysis method based on the least square method to quantitatively determine a reasonable upward-continuation height. The steps of this method are as follows. Perform upward continuation to different adjacent heights for observation data, and then use the least square method to estimate the least square error of the upward continued value of adjacent heights.There is a maximum curvature in the least square curve of upward-continued values of all adjacent heights.At the point of the maximum curvature, the local anomalies are attenuated to the greatest extent, while the regional anomalies are preserved as far as possible. Therefore, this point can be approximately regarded as the optimal upward-continuation height. As indicated by tests using the data of a theoretical model, the method proposed in this paper can be used to qualitatively determine a suitable upward-continuation height, thus providing an important reference for the selection of upward-continuation height in practical applications.