The article presents a statistical model of dynamic viscosity changes in engine oils. The dynamic viscosity changes at the temperature of 150°C and a shear rate of 129 and 258 s-1 were calculated for five in-service oil groups tested. A detailed description of oil groups and the car fleet was presented in the first part of the article series. The aim of the study was to develop a statistical model based on dynamic viscosity changes which could be used to predict oil condition changes during operation. The models of linear regression were prepared on the basis of the least squares method. The estimated regression equation in the form of Y = A + B*X, makes it possible to calculate average predictive values for parameters (Y) at given mileage (X). Although the oils belonged to the same viscosity class and were operated under similar conditions, engine oils from the PE group had a significantly lower dynamic viscosity compared with other oil groups. They were also characterized by the least steep slope of the dynamic viscosity curve and the obtained values remained within the limits of approximately 45,000 km (extrapolation). Prediction intervals were also relatively narrow in this group of oils, which is confirmed by the lowest standard error of the estimate among the five oil groups analyzed. Point predictions for the engine oils tested at two shear rates indicate that there are close correlations between the two. The obtained results may facilitate decision-making regarding oil change schedules.