This paper presents a method for extracting transient events from the power system signal based on independent component analysis. The extraction of transients is important to diagnose the system in terms of power quality, because the computation of power quality indices from the extracted events is more efficient and easier to be carried out. Furthermore, the proposed method can be applied as a preprocessing tool for transient characterization in disturbance detection, classification, and identification systems. Both qualitative and quantitative analyses demonstrate the efficiency and the advantages of the proposed method in comparison with other approaches commonly used for the same purpose.