Chitosan has been proposed for a variety of tissue engineering scaffold. In this study, we evaluated the in vivo biocompatibility of chitosan scaffolds with various deacetylated degree(DD) in rat, mouse and rabbit models. The structure of the scaffold was improved by addition of butanol during the lyophilization(-B) and the DD of the chitosan scaffold(CS) was modified by treatment with acetic anhydride to fully acetylated form(ACS) or with NaOH to fully deacetylated form(AFCS). To evaluate biocompatibility of chitosan with different DD, the scaffolds were implanted subcutaneously, and animals sacrificed after 3, 15, 30, and 60 days in rats and after 1, 3, and 9 weeks in mice and rabbits. In rat implantation models, the improved biocompatibility of the ACS was demonstrated by matrix-rich neo-tissue formation with much less infiltration of polymorphonuclear cells(PMNs) and faster degradation after subcutaneous implantation compared to CS scaffolds. However, in mouse and rabbit implantation models, the severe immune reaction of the CS-B was demonstrated by large numbers of infiltration of neutrophils during 9 weeks and the very slow degradation after subcutaneous implantation of ACS-B and AFCS-B was observed. To evaluate the effectiveness of chitosan scaffold as a dermal substitute, the scaffolds were implanted onto the full thickness skin wound in rat and rabbit models. The ability of chitosan scaffold to repair full thickness dermal defects was evaluated by histological examination of transplanted skin. In rat skin wound model, transplantation of ACS as well as CS restored dermis, in contrast, the severe inflammatory reaction that occurred in rabbit skin wound model. In conclusion, we showed that the biocompatibility of chitosan scaffolds can be variable depending on the DDs as well as on animal species. Therefore, more comprehensive comparative studies may be necessary before the clinical application of the chitosan-based scaffold.