Methods Sandblasted CAD/CAM composite blocks (LAVA Ultimate, 3M ESPE) were cemented to bur-cut dentin using either the etch & rinse composite cement Nexus 3 (‘NX3’, Kerr) with Optibond XTR (‘XTR’, Kerr), or the self-etch composite cement RelyX Ultimate (‘RXU’, 3M ESPE) with Scotchbond Universal (‘SBU’, 3M ESPE). All experimental groups included different ‘curing modes’ (light-curing of adhesive and cement (‘LL’), light-curing of adhesive and auto-cure of cement (‘LA’), co-cure of adhesive through light-curing of cement (‘AL’), or complete auto-cure (‘AA’)) and different ‘restoration-surface pre-treatments’ of the composite block (NX3: either a silane primer (Kerr), or the XTR adhesive; RXU: either silane primer (RelyX Ceramic Primer, 3M ESPE) and SBU, or solely SBU). After water-storage (7 days, 37°C), the μTBS was measured. Additionally, the degree of conversion (DC) of both cements was measured after 10min and after 1 week, either auto-cured (21°C/37°C) or light-cured (directly/through 3-mm CAD/CAM composite).ResultsThe linear mixed-effects model (α=0.05) revealed a significant influence of the factors ‘curing mode’ and ‘composite cement’, and a less significant effect of the factor ‘restoration-surface pre-treatment’. Light-curing ‘LL’ revealed the highest μTBS, which decreased significantly for all other curing modes. For curing modes ‘AA’ and ‘AL’, the lowest μTBS and a high percentage of pre-testing failures were reported. Overall, DC increased with light-curing and incubation time.SignificanceThe curing mode is decisive for the bonding effectiveness of adhesively luted composite CAD/CAM restorations to dentin.