OBJECTIVESTo compare mineral density of residual dentine after excavation with different caries-removal techniques and to evaluate the diagnostic potential of laser-induced fluorescence (LIF), measured by DIAGNOdent, as a tool to determine the caries-removal endpoint.METHODSCarious teeth were excavated by tungsten-carbide round burs (Komet), ceramic burs (CeraBurs, Komet), sono-abrasion (Cariex TC tips, Kavo), and by chemo-mechanical excavation using two enzyme-based solutions (exp. SFC-V and SFC-VIII, 3M-ESPE) or a sodium hypochlorite-based solution (Carisolv, MediTeam). The caries-excavated teeth were scanned by micro-CT (1172, Skyscan), after which the mineral density at the bottom dentine was correlated to LIF measurements at the same region. A micro-CT threshold for dentine caries was defined by comparison with surface-hardness measurements. The intensity of dentine staining was evaluated by analysing the component 'L*' in CIE-L*a*b-converted images from the excavated teeth.RESULTSNo statistically significant difference in mineral density was found at the bottom of the cavities prepared with the different caries-excavation techniques, except for exp. SFC-V that left residual dentine with a significantly higher mineral density than when CeraBurs were used (Tukey-Kramer, p<0.05). Absence of residual caries was associated with darker staining of dentine. No significant correlation was found between the distance from the deepest cavity point to the pulp-chamber roof and LIF measurements. A strong negative correlation (R=-0.86, p<0.01) was however found between L* values and LIF measurements, indicating that staining in residual dentine leads to higher LIF measurements.CONCLUSIONSLIF measured by DIAGNOdent is influenced by staining in residual dentine. Therefore, its use to determine the caries-removal endpoint is doubtful.