Objective: The aim of this in vitro study was to compare the performance of a laser fluorescence device (DIAGNOdent) and a caries-detector dye for detecting residual dentinal caries. Methods and Materials: Forty extracted human molar teeth with deep dentinal carious lesions on their approximal surfaces were used for the study. The caries were removed by a new chemomechanical system, Carisolv. Three independent observers checked all cavities for residual caries using caries-detector dye and DIAGNOdent. The presence or absence of residual caries was verified using polarized light microscopy as the gold standard. Three observers recorded two different measurements at the same site after 2 weeks to determine intra- and interexaminer reproducibility of the DIAGNOdent. Results: Average kappa values of 0.61 for interexaminer reproducibility and 0.94 for intraexaminer reproducibility were determined. The specificity of caries-detector dye was found to be 100% and the specificity of the DIAGNOdent was 86%. Conclusion: Although DIAGNOdent showed greater sensitivity than caries-detector dye (P .05), low sensitivity values for each diagnostic system may lead to teeth with residual caries being undetected. Therefore, it might be advisable to reconsider the criteria for detecting residual caries during tooth preparation. Keywords: caries-detector dye, Carisolv, DIAGNOdent, residual dentinal caries