Abstract: Different wavelet bases have different time-frequency characteristics.Different results will come out while analyzing the same problem with different wavelet bases.So the selection and the optimization of wavelet base are important in the application of wavelet analysis.In this paper basic features of wavelet base are expounded and the correlation coefficient is regarded as the basis of selecting optimum wavelet base;four wavelet base functions are chosen to be analyzed respectively and the optimum wavelet base of multiscale analysis in well logging is determined by analyzing the magnitudes of correlation coefficients.Reconstructing the original logging curves using multiscale anlysis shows that the selection of optimum wavelet base is reasonable and the correlation coefficient as a method of selecting wavelet base is feasible.