Abstract: Based on the theory of wavelet transform,a new method by using wavelet transform to eliminate the noise in observed data is introduced and a calculation example is also given. The results show that denoise using wavelet transform is efficient and reliable and is particularly suitable for processing monitoring data of building deformation due to its sensitivity to the noise of data and the ability to identify useful information. At the same time it is also indicated that the key problem of obtaining good analytical results is to choose a proper wavelet function,and the decision of denoise threshold is also very important. It is proved by the results that using db3 wavelet function and soft threshold in the denoising of monitoring data of building deformation is appropriate and wavelet transform provides a new and efficient method in this field. Some problems worth noticing in practical application are briefly explained.