Abstract. This contribution presents a system, which allows for a continuous monitoring of speech intelligibility from a single microphone signal. The system accounts for the detrimental effects of environmental noise and reverberation by estimating the two relevant parameters signalto-noise ratio and reverberation time, and feeding them to a speech intelligibility model. Due to its real-time functionality and the fact that no reference signal is required, the system offers a wide range of opportunities to monitor communication channels and control further signal enhancement mechanisms. A priori knowledge of the individual hearing loss can be used to make the system applicable also for hearing-impaired users.