Fold-switching enables metamorphic proteins to reversibly interconvert between two highly dissimilar native states to regulate their protein functions. While about 100 proteins have been identified to undergo fold-switching, unveiling the key residues behind this mechanism for each protein remains challenging. Reasoning that fold-switching in proteins is driven by dynamic changes in local energetic frustration, we combined fold-switching simulations generated using simplified structure-based models with frustration analysis to identify key residues involved in this process based on the change in the density of minimally frustrated contacts during refolding. Using this approach to analyze the fold-switch of the bacterial transcription factor RfaH, we identified 20 residues that significantly change their frustration during its fold-switch, some of which have been experimentally and computationally reported in previous works. Our approach, which we developed as an additional module for the FrustratometeR package, highlights the role of local frustration dynamics in protein fold-switching and offers a robust tool to enhance our understanding of other proteins with significant conformational shifts.