The risk factors associated with road accident are directly related to the characteristics of the roadway, the vehicle type and the behavior of the driver, among others. For that reason, such traffic elements are intensively investigated and analysed in the field of road safety. Among the techniques and methods developed, those based on statistical analysis have demonstrated a high degree of susceptibility to the problem and have been applied in several studies of traffic accidents. In this perspective, this work presents a theoretical and applied discussion of the technique of Principal Component Analysis (PCA), in the study of road accidents. The main objective is to contribute to the discussion and theoretical foundation of the statistical techniques, used in the multivariate analysis of highway databases, generated by roadway concessionaries. The database used for this study is from the Dom Pedro I Highway, located in the urban area of the Campinas city in Brazil, during the period of four years from 2009 to 2012.