Aims: Bioactive glasses are a family of silica-based synthetic biomaterials, which form chemical bonding with the surrounding bone. The limiting biologic factors of the bonding process are poorly understood. The hypothesis of the current study was that there are species-specific differences in the incorporation of bioactive glasses due to anatomic and physiologic factors. Methods: Conical porous implants made of sintered bioactive glass or titanium microspheres (Ø 250–300 μm) were surgically implanted bilaterally into the cortex of tibias or femurs in sheep, dog and rabbit. Implant incorporation was evaluated by means of push-out testing, pQCT, his-tomorphometry, BEI-SEM, and EDXA. The comparison was made at 12 weeks. A total of 176 implants were analysed. Results: Between the three species, there were significant differences in the extent of new bone ingrowth and in the mechanical strength of implant fixation. The rabbit showed the highest amount of bone ingrowth into both bioactive glass and titanium implants. Also the shear strength of the implants was superior in the rabbit compared with the dog and the sheep. Histological pattern of new bone ingrowth into bioactive glass structures was similar in the dog and in the rabbit. In contrast, the ingrowth of new bone failed into bioactive glass implants in the sheep. Conclusions: Based on these results, the sheep represents a divergent model for bone healing studies of bioactive glass. Long bones of the sheep contain yellow (fat) marrow and we assume that the poor healing response reflects the deficiency of marrow-derived osteoprogenitor cells.