[en] Although the information provided by the evolution of the density matrix of a quantum system is equivalent with the knowledge of all observables at a given time, it turns out ot be insufficient to answer certain questions in quantum optics or linear response theory where the commutator of certain observables at different space-time points is needed. In this doctoral thesis we prove the existence of density matrices for common probabilities at multiple times and discuss their properties and their characterization independent of a special representation. We start with a compilation of definitions and properties of classical common probabilities and correlation functions. In the second chapter we give the definition of a quantum mechanical Markov process and derive the properties of propagators, generators and conditional probabilities as well as their mutual relations. The third chapter is devoted to a treatment of quantum mechanical systems in thermal equilibrium for which the principle of detailed balance holds as a consequence of microreversibility. We work out the symmetry properties of the two-sided correlation functions which turn out to be analogous to those in classical processes. In the final chapter we use the Gaussian behavior of the stationary correlation function of an oscillator and determine a class of Markov processes which are characterized by dissipative Lionville operators. We succeed in obtaining the canonical representation in a purely algebraic way by means of similarity transformations. Starting from this representation it is particularly easy to calculate the propagator and the correlation function. (HJ) 891 HJ/HJ 892 MKO