In this paper Kaiman filter and Wavelet transform (DWT) are compared in voltage event segmentation. The waveform segmentation is required to detect stationary and non-stationary stages throughout the voltage event waveforms. This detection plays an important role in the features extraction process since some features must be computed during stationary stages and other in non-stationary ones. The comparison is carried out using field measurements, which have been previously segmented through a visual inspection. The third and fourth states of the Kaiman filter as well as the voltage fundamental component have been used in the waveform segmentation process. Kaiman- and DWT-based segmentation results are compared and their advantages and drawbacks are discussed.