BackgroundThis in vitro study aimed to assess carious lesions on root surfaces using quantitative light-induced fluorescence (QLF) and to compare the readings with axial lesion depth on µCT. MethodsThe root surfaces of 107 extracted human teeth were included after visual-tactile inspection. For further analysis, the following parameters were assessed: clinical findings (non-cavitated: leathery or hard, cavitated), QLF- (QLF-D Biluminator 2+), and µCT-images (Bruker Skyscan 1172). The shape of the undamaged tooth surface of the cavitated lesions was virtually re-constructed during µCT analysis. Clinical surface texture,% fluorescence loss, and lesion depth (µCT) were determined. Statistical analysis: chi²-test, Spearman-Rho test, regression analysis. Results∆F was significantly lower in non-cavitated leathery (-50.37 ± 15.10) and cavitated (-61.23 ± 9.92) compared to non-cavitated surfaces with a hard texture (-17.04 ± 16.10, p < 0.01). For non-cavitated surfaces, a negative correlation was observed between ∆F and lesion depth in µCT images regardless of texture (-0.748, p < 0.01). Regression analysis revealed that ∆F predicted lesion depth in µCT for non-cavitated surfaces (β: 0.703, CI95: 0.67–-0.43, p < 0.01). ConclusionThe percentage of fluorescence loss (∆F) in QLF predicted lesion depth of non-cavitated demineralized root surfaces. Therefore, QLF can be recommended for estimating the lesion depth of carious root lesions and seems to expand the possibilities of follow-up and lesion monitoring, especially for non-cavitated surfaces.