The proposed research will, by exploring the relationship between process algebras and Markov chains, develop efficient techniques which may be applied in a systematic way to the performance analysis of large, complex systems. On the practical side, we will utilise the resulting methodology to address the modelling and analytical challenges presented by the next generation of computer and communication systems. Markovian process algebras (MPA) are extensions of classical process in which time and probabilistic elements are introduced. The major advantage of MPAs is their inherent compositional structure which facilitates hierarchical construction and analysis of models. The proposed research will exploit the benefits of the MPA approach by using particular structures identified within models to enhance model solution.