The use of dog teeth in vivo provides a useful screening model between in vitro tests and human clinical trials. In this experiment, the effects of post-insertion time on the structure and strength of resin-dentin bonds were evaluated. The purpose of this experiment was to determine the resin-dentin bond strengths in dog teeth in vitro vs. in vivo ( immediate, 3 weeks and 3 months after restoration ), and to investigate the resin-dentin interfaces with scanning electron microscopy ( SEM ). Box-form Class V cavities were prepared into labial middle coronal dentin in 4 anesthetized beagle dogs. The teeth were bonded with Fluoro Bond ( FB, Shofu ) or Clearfil Liner Bond 2 ( LB, Kuraray ) and restored with Clearfil AP-X resin composite ( Kuraray ). The teeth were surgically removed immediately, 3 weeks or 3 months after restoration, serially sliced and trimmed to a surface area of 1mm² for microtensile bond strength test ( MTBS ). For the in vitro study, extracted dog teeth were used. In vitro and in vivo immediate bond strengths of both adhesive systems were 33.7 to 36.3 MPa, which were not significantly different. In vivo bond strengths of both adhesive systems at 3 weeks and 3 months were significantly lower than those measured immediately in vitro or in vivo. SEM showed the formation of hybrid layers with both adhesive systems when tested immediately in vitro or in vivo, but after 3 month in vivo, hybrid layers could not clearly seen. The fall in bond strengths noted with time may have been due to occlusal stress, and the influence of oral or dentinal fluids on the morphology of the hybrid layers and resin tags.