The invention discloses a denoising method of microseismic signals in a well based on the 3D shearlet transform. The method comprises the following steps of: step 1, converting three component data ofthe microseism in the well into a three-dimensional data set; step 2, performing shear transformation on the data set to obtain a shear coefficient; step 3, performing threshold processing on the shear coefficient to obtain a post-threshold coefficient; step 4, applying a reverse shear transformation to the post-threshold coefficient to obtain a time domain denoised signal; and step 5, outputtingthe first three surfaces of the time domain denoised signal to obtain a valid signal. The denoising method of microseismic signals in the well based on 3D shearlet transform utilizes the correlationbetween the microseismic data and the coefficient distribution in the shear domain, applies the 3D shearlet to process the three component signals, and uses the multi-scale threshold function for denoising, thus achieving excellent performance in terms of effective signal retention and noise attenuation.