Perfusion processes have gained interest as mammalian cell culture mode. Due to the high complex issues and variability regarding such culture mode, a Principal Component Analysis (PCA) to the data was used to characterize such variable process patterns and to achieve a better understanding. The transfected NS0/1B8 cell line was fermented in a 500 L bioreactor in perfusion culture mode to obtain a desired monoclonal antibody against CD20 molecule. Given the high variability of the process, an exploratory data analysis based on a multivariate analysis technique such as PCA was applied. The variables were selected by a risk model analysis based on the cause-effect matrix, the experience accumulated in the process, over the critical quality attributes, and parameters focus on the fermentation process. As a result, it was obtained that two main components were able to explain more than 95% of the total variance, and it was possible to select between the critical parameters those that have the greatest contribution to the variability of the fermentation process. Furthermore, the practical experiences of the specialists matched with the results and new process recommendations were projected to improve the control strategy for a further Continuous Process Verification. Keywords: Monoclonal antibody, Principal Component Analysis, Fermentation, Critical parameters.