A combined wavelet analysis - Kalman filtering approach is presented in this paper for the automatic detection and analysis of voltage dips in power systems. Wavelet analysis is used to process the voltage waveform for detection and better estimation of the time-related parameters of a voltage dip in voltage supply, while Kalman filtering is used for the estimation of voltage magnitude and phase angle during the event. The method proposed, which can be processed in real-time, has been tested both by computer simulations and experimentally using measurements from a low voltage distribution system, showing the potential of this combined scheme for complete and accurate characterization of voltage dips in power systems.