The theory of Markov chains is used to develop a procedure for deriving a general dimensionless parameter that allows prediction of the dynamic behaviour of continuous-flow chemical systems. The basis of this procedure is the Markovian intensities defined for all single-transport processes (as convection, diffusion, adsorption, reaction, etc.) occurring in the system under consideration. The corresponding parameter is derived by combining the Markovian intensities according to a modular construction principle which rests upon utilizing certain basic structures. With this principle the parameter may be obtained for complex systems, i.e. for systems with a large number of single physicochemical transport and conversion processes. The parameter quantitatively determines the deviation of the actual process from the corresponding equilibrium process. Linear as well as strongly non-linear systems can be described without time-consuming numerical simulation procedures. The proposed procedure is successfully used to predict the dynamics of two isothermal non-linear adsorber models.