Summary. We are considering molecular dynamics and (sequential) membrane systems from the viewpoint of Markov chain theory. The first step is to understand the structure of the configuration space, with respect to communicating classes. Instead of a reachability analysis by traditional methods, we use the explicit monoidal structure of this space with respect to rule applications. This leads to the notion of precycle, which is an element of the integer kernel of the stoichiometric matrix. The generators of the set of precycles can be effectively computed by an incremental algorithm due to Contejean and Devie. To arrive at a characterization of cycles, we introduce the notion of defect, which is a set of geometric constraints on a configuration to allow a precycle to be enabled, that is, be a cycle. An important open problem is the efficient calculation of the defects. We also discuss aspects of asymptotic behavior and connectivity, as well as give a biological example, showing the usefulness of the method for model checking. Corresponding author186 M. Muskulus et al.