This eaper aims to show how multivariate statistical procedures, principal component analysis (PCA) and correspondence analysis (CA), can be used to identify sorne critical mineral crystallization factors during metamorphism of Mn-rich lithologies. The data set was obtained from a petrographic and geochemical study of Mn-lithologies from Southern Iberian Massif, Spain. We considered three types of variables in the system: mineralogical variables, bulk chemistry variables and a physical variable (oxygen fugacity). PCA was applied to the compositional variables with the aim of recognizing groups of elements according to the geological evidence. We obtained four components, which were used as new variables that account for the variance of the compositional data in a more summarized manner. These new compositional variables, together with the mineralogical and physical variables, were treated in the CA. The plot of the variables on the factor l-factor 2 plane reveals that tephroite crystallization is controlled by a compositional variable representing low values of ratio Si/(Mn + Fe) in the rocks. Moreover, it is possible to deduce from this plot the importance of high oxygen fugacity conditions on the crystallization of piemontite. The third factor shows how the crystallization of spessartine does not require very restrictive physical-chemical conditions.