A method of image noise reduction based on multiwavelet transform and data fusion is proposed. With this method, several images including same original signal and different noises are decomposed by wavelet transform. In high frequency area, the coefficients are processed by threshold operation, and then important wavelet coefficients are selected according to multiscale data fusion. In low frequency, the new approximation coefficients are obtained directly by the weighted mean value of the coefficients in different images. In addition, the fused image can be obtained by using the inverse multiwavelet transform for all important coefficients and approximation coefficients. Experimental results show, using this method, image noise can be reduced effectively and little image detail is lost.