The role of vertical gradient of gravlty anomaly in the solution and interpretation of geological and geophysical problems on superficial layer is increasingly evident. However, it is difficult to directly observe the vertical gradient of gravity anomaly in mountainous area. Moreover, aerial and satellite gravity gradient measurments have not been put into practice up to date yet. The vertical gradient of gravity anomaly is calculated by use of topographic data and several problems concerned are investigated and discussed. It turns out in theory and practice that the relative error of the vertical gradient determined with the method is only 5%. and when the vertical gradient of gravrty anomaly reaches the maximum 1 000 ns−2, the error is only 50 ns−2. In addition, how to divide ring zones in calculation and the errors caused by the inaccuracy of landform elevation and rock density has been quantitatively analysed.