Multivariate process control charts have been increasingly popular to monitor many different industrial processes. There are many reasons for this, but the main one is the recent advances that have occurred in multivariate quality control methods. For processes where huge amounts of multidimensional data are available, multivariate projection methods, such as principal component analysis (PCA) and partial least squares (PLS), have received much attention by a variety of industries. This paper introduces a new technique that can be used in the early process monitoring design stage to select from among the set of quality characteristics or process parameters a smaller set that is adequate to process control and ensure a product satisfies yield specifications.