Voice over the Internet Protocol (VoIP) has been increasingly popular, but reliability and voice quality remain important factors that limit the widespread adoption of VoIP systems. Providing good voice quality is of major importance for the transition from the PSTN to VoIP networks. There are several non-real-time algorithms that estimate the voice quality such as the PESQ and the E-model. In this thesis we propose a real-time fuzzy algorithm to estimate the echo quality component of the voice quality in VoIP networks. Differently from the existing algorithms, the proposed algorithm does not need a reference signal and has low computational complexity. For these reasons, the proposed algorithm can be embedded in every VoIP system of a network to monitor live calls, giving an estimate of the instantaneous voice quality to the network provider.