The aim of this study was to determine the effect of the rotor speed, the amount of water sprayed and the atomizing air pressure on the size, size distribution and shape of the pellets produced in a fluid bed rotor granulator, using microcrystalline cellulose and α-lactose monohydrate as starting materials. It is a follow up study based on the conclusions of a 3 3 full factorial design performed previously. Initially, a new 2 3 full factorial design with center points was set up. However, the results of the analysis of variance, revealed the lack of fit of the linear model for the size and shape factors. Therefore, the original 2 3 factorial design was expanded to a central composite design in order to fit a second order polynomial that was then found to describe the data adequately. An increase of the amount of water added, resulted in larger pellets with increased sphericity, while an increase of the rotor speed decreased the size of the pellets and increased their sphericity. All the above-mentioned effects were found to be statistically significant (p < 0.05). The effect of the atomizing air pressure on the size of the pellets, their size distribution and their shape was not statistically significant. Significant interactions between the factors for the size and the shape were also found. The propagation of error and desirability function were utilized to identify the levels of the selected factors that simultaneously optimize all the responses while reducing the variation transmitted to the responses from the variability of the controlled factors. Finally, further experimental runs were performed to confirm the validity of the obtained equations that describe the relationship between the factors under study and the responses.