The basic idea and the advantage of wavelet analysis are described, and the index of the summer drought intensity in east of the Yellow River in Gansu Province from 1951 to 1995 are calculated using Mexican cap wavelet and Haar wavelet analysis methods, respectively. The results show that the summer drought events in east of the Yellow River in Gansu Province is composed of oscillations on different time scales. In the middle of Guansu Provine, there is an obvious oscillation period of quasi-19 years, with two ascending and two descending episodes. In east and south Guansu, there are obvious oscillation periods of quasi-22 and quas-10 years, respectively, with one ascending and two descending episodes. Reconstruction of original time series of summer drought intensity index in terms of different time scales indicates that the wavelet analysis method could be used in filtering.