Data replication is one of the main techniques by which database systems can hope to meet the stringent temporal constraints of current time-critical applications, especially Web-based directory and electronic commerce services. A pre-requisite for realizing the benefits of replication, however, is the development of high-performance concurrency control mechanisms. We present in this paper MIRROR (Managing Isolation in Replicated Real-time Object Repositories), a concurrency control protocol specifically designed for firm-deadline applications operating on replicated real-time databases. MIRROR augments the optimistic two-phase locking (O2PL) algorithm developed for non real-time databases with a novel and simple to implement state-based conflict resolution mechanism to fine-tune real-time performance.Using a detailed simulation model, we compare MIRROR's performance against the real-time versions of a representative set of classical protocols for a range of transaction workloads and system configurations. Our performance studies show that (a) the relative performance characteristics of replica concurrency control algorithms in the real-time environment could be significantly different from their performance in a traditional (non-real-time) database system, (b) MIRROR provides the best performance in both fully and partially replicated environments for real-time applications with low to moderate update frequencies, and (c) MIRROR's conflict resolution mechanism works almost as well as more sophisticated (and difficult to implement) strategies.