Distributed real-time systems based on local area networks are becoming increasingly common. Communication between activities executed on different nodes open have strict timing constraints.As new proposals for time-critical protocols continue to emerge, a need to provide a uniform set of pegormance criteria to enable potential users to make accurate and objective comparisons between different solutions, has become apparent. The multiple access protocols which have been standardised to date were primarily intended for general data traffic. The three most commonly used peflormance measures -information throughput, channel utilisation, and mean transfer delay -are not sufficient to determine whether a given protocol will meet the speciJic timing requirements. In this paper we identifr the issues encountered and strategies used when testing the peqormunce of time-critical protocols.