This thesis presents results about the development of a tool for real time monitoring of the voltage stability margin of transmission systems. The advance of Phasor Measurement Units (PMU), has made it easier to obtain the system's state, voltage magnitude and phase angles of some buses in real time. The access of the accurate system's state in real time allows the utilization of parametric system identification techniques to estimate the voltage stability margin. In this thesis linear and nonlinear Auto-Regressive eXogenous inputs (ARX) models were calculated based on an offline database and utilized to represent the relation between the PMU measures and the voltage stability margin. Two criteria for the allocation of the PMUs in the network were tested: the voltage variance and the modal participation factor. The results demonstrate that the proposed method presents adequate accuracy for the estimation of the voltage stability margin even considering the occurrence of contingencies. The estimation is performed through the execution of relatively simple operations on the measured values instead of computing several PV curves, which makes it suitable for real time applications.