The surface topography often is an important quality criterion for the manufacturing of milled workpieces as it often defines their functional behaviour. In machining both, the kinematics of the process and the stochastic influences deriving from the machine tool, workpiece and the surrounding environment affect the workpiece's surface roughness. This paper presents a simulation-based method for flank milling, which considers kinematic and stochastic influences including run-out errors and tooth length variations. The simulation results are used in combination to predict the surface roughness depending on the chosen process parameters. Hence, also making in possible to choose appropriate process parameters to achieve a defined surface roughness.