Abstract-This paper deals with an efficient methodology aimed at monitoring the radiated electromagnetic emissions along a high-speed railway system in the hundreds of kilohertz range. In particular, the proposed approach allows a compressed representation of the spatial distribution of the frequency spectrum of the radiated magnetic field generated by the currents placed on the railway conductors by electrical apparatus on board of running railway vehicles. The main idea underlying this work is that the standing wave nature of current distribution along the railway line results in a spatial distribution of radiated magnetic field which can be effectively represented by resorting to the emerging compressive sensing theory. To this aim, wireless magnetic-field sensors are assumed to be deployed along the railway line and used to provide spatial samples of the magnetic field spectrum. The main advantages of the proposed approach include a smaller number of sensors when compared with the number foreseen by the straightforward use of the conventional Nyquist-Shannon sampling approach, and a simple treatment of nonuniform spatial distribution of sensors. Suitability of the proposed approach is supported by measurement data and electromagnetic models already available in the related literature, whereas effectiveness of field spatial reconstruction is proved through numerical simulations. Although the application presented in this work is specific to the magnetic field distribution in a limited frequency range, the proposed approach has a general validity and could be effectively exploited for distributed monitoring of other physical quantities, in other frequency ranges, related to electromagnetic compatibility and safety/security issues in high-speed railway systems.