In this chapter, we critically examine the application of global sensitivity analysis (GSA) techniques for variable screening in evolutionary optimization, particularly focusing on their effectiveness in reducing dimensionality for problems optimized via CMA-ES. We assess the use of the GSAreport tool, which generates sensitivity reports aimed at identifying what parameters are influential on a target of a given problem and their interactions. We analyze two test cases from engineering design: (1) a star-shaped crash box subjected to axial impact, evaluating crashworthiness, and (2) a sheet metal forming application, assessing the drawability of metal components. While GSA methods can identify variables with significant impact on model outputs, we find that relying solely on GSA for dimensionality reduction may limit optimization effectiveness when using adaptive algorithms like CMA-ES. In these case studies, we demonstrate that while GSA aids in understanding variable importance, the exclusion of “less influential” variables may oversimplify the problem space, potentially hindering CMA-ES’s capacity to explore and adapt to complex landscapes.