: Refactoring is a process meant to improve the internal quality of software systems. However, while on one hand, the guidelines for this delicate process are still empirical and qualitative, on the other hand, software product metrics often indicate that this process has the opposite results. Also, there is a lack of evidence regarding improvements on maintainability due to refactoring. This means that this process, although widely acknowledged as one of the best software practices, is difficult to deploy within large scale software systems, and can be better grounded. To address these challenges, we propose a method for refactoring with quantitative and experimental grounds. Upon the consolidation of this method, we will build the necessary blocks to implement and validate it.