Pulse wave is easily contaminated by noise when being sampled and processed. In order to extract and analyze the useful information involved, the paper presents a novel denoising algorithm based on empirical mode decomposition (EMD) and independent component analysis (ICA).First, the signal is decomposed by EMD ,then a series of intrinsic mode functions (IMFs) and the remainder can be obtained. Second, ICA will be performed to separate independent components from IMFs. At last, these independent components will be used to reconstruct the pulse wave signal. By using the method, we are able to separate noise and estimate desired source signals from pulse wave signal. Experimental results substantiate the strong potential of the proposed method for pulse wave signal noise separation.