A series of multivariable models composed from linear differential equations, integrals, and delays has been developed to describe the dynamics of mixing of the raw materials during the cement raw meal milling. The dynamics between the main oxides and chemical modules of raw mix in mill outlet and the materials proportions in mill entrance is investigated. The distributions of model parameters (gains, time constants, delays, steady state values) are determined. In this way, the structure of the parameters' uncertainty is clearly expressed, explained by the raw materials composition uncertainty and can be used in process simulation. A successful attempt is made to simulate the mill operation by including the main components of the process concerning mixing and dynamics. The simulator calculates the chemical analyses and modules of raw meal in the mill outlet, and also the dynamics of the main module characterizing raw meal and cement i.e., the lime saturation factor (LSF). Thus, the effect of the raw materials compositions and their uncertainty on LSF dynamics is studied. The dynamical models analyzed can be used to build robust controllers, while the simulator constitutes a tool to evaluate the efficiency of the control strategies.