Aim: During removal of resin-based composite (RBC) restorations, removal of adjacent sound tooth structure can easily occur. The aim of the study was to compare the fluorescence-aided identification of restorations (FAIR) method with the conventional method (CM) using white light illumination for the selective removal of tooth-colored RBC restorations. Materials and Methods: Sixty extracted teeth were used to prepare 15 sets of models, each with four teeth. Cavities were prepared in each tooth and a digital three-dimensional scan of each model was taken in STL format. The teeth were then restored with either Admira Fusion, GRADIA DIRECT X, or TPH Spectra LV (20 teeth for each). Five dentists removed the restorations from sets of 12 teeth (six each using FAIR, and six using CM). From a postoperative scan, changes in intercuspal cavity width were measured using the 3D tool CAD software. Results: Using FAIR, changes in cavity width were not significant (P =.17), and there was no significant interoperator variance (P =.3). In contrast, when using CM, there was a significant increase in cavity width (P =.0025), and considerable interoperator variance (P =.03). FAIR allowed faster completion of the task of restoration removal (average time 100.23 seconds, vs 165.13 seconds from CM). Conclusion: The FAIR method facilitates the selective removal of tooth-colored RBC restorations, with reduced time and less inadvertent removal of sound tooth structure. Clinical Significance: By enhancing the differentiation between the restorative material and adjacent tooth structure, the FAIR method facilitates removal of tooth-colored restorations, while concurrently prevents unnecessary removal of healthy tooth structure. The same approach may have applications for removal of resin-based materials such as resin cements under restorations, resin fissure sealants, and orthodontic brackets bonding resins.