Objectives: The objective of this study was to study the effect of different adhesive layers on the interfacial fracture toughness (K-ICi) of dentin-resin composite interfaces. Methods: Miniature short-rod fracture toughness specimens containing a chevron-shaped dentin-bonded interface along their midplane were used for testing. Each interface zone contained a thinned (one coat of unfilled adhesive resin, air-thinned), one-layer (one coat of unfilled adhesive resin, brush-thinned), two-layer (two coats of unfilled adhesive resin, brush-thinned), 10% filled or 45% filled adhesive resin layer. After storage in distilled water at 37 degreesC for 24 h, the fracture toughness test specimens were loaded in tension at an extension rate of 0.5 mm/min until fracture and the K-ICi were determined. The results were analysed by ANOVA and Fisher's LSD test (p < 0.05). Scanning electron microscopy was used to examine representative fracture surfaces. Results: There were no significant differences in mean K-ICi among the different unfilled adhesive resin layer groups. SEM examination of these specimens showed that fracture generally occurred between the resin-infiltrated layer and adhesive resin layer during interfacial fracture toughness testing. The mean K-ICi for the 10% filled groups was not significantly different from the unfilled groups. The 45% filled group, however, demonstrated the highest K-ICi values, the thickest adhesive resin layer under SEM examination, and a fracture path through the adhesive resin layer. Conclusions: There were no significant differences in K-ICi when the unfilled adhesive resin was used despite different application methods. The 45% filled adhesive resin improved the properties of the dentin-composite interface with respect to both interfacial fracture resistance and dentinal seal after fracture. (C) 2001 Elsevier Science Ltd. All rights reserved.