Aims: This study uses radiological and clinical evaluations of the healing of endosseous titanium implants presented with different surface characteristics in the clinical situation (paper I-III) and experimentally to describe the early bone healing in maxillary sinus membrane elevation with and without the use of grafting material (paper IV). Material and methods: In paper I, 136 patients were treated with 394 dental implants – 199 were oxidized titanium implants (Nobel Biocare TiUnite) and 195 were turned titanium surface implants (Nobel Biocare Mark III). Implant survival rates were retrospectively investigated after a minimum of five months after functional loading of the implants. At the five-year follow-up (paper II), eight patients were deceased and 128 were invited. Twenty-five patients refrained from participating in the study. The remaining 103 patients (287 implants – 133 with a turned surface and 154 with an oxidized surface) were examined after at least five years of functional loading. Clinical examinations of bleeding on probing (BoP) and pocket depth (PD) were performed. Intraoral radiographs were used to assess marginal bone levels (MBLs). In paper III, 28 patients were subjected to autologous bone graft and delayed implant placement, with a total of 92 dental implants. Thirteen patients received 47 implants with a turned surface and 15 patients received 45 implants with an oxidized surface. After a minimum of five years of functional loading, all patients were clinically examined regarding PD and BoP. The MBL was measured in intraoral radiographs. Cone beam computed tomography (CBCT) was used to evaluate the apical bone level (ABL) of the implants and intra-sinus conditions. The experimental study (paper IV) used nine adult male tufted capuchin primates (Cebus apella). Eight animals were subjected to bilateral maxillary sinus membrane elevation using a lateral replaceable bone window technique. One oxidized dental implant was placed in the residual bone of the sinus floor, protruding into the maxillary sinus cavity on both sides. In four animals, one sinus was left without any additional treatment, while the contralateral sinus was filled with autologous bone grafts from the tibia. In two animals, the implants were inserted under the elevated sinus membrane on both sides. In two animals, the sinus membrane was totally removed bilaterally before placement of implants. The animals were euthanized after 10 (n=4) or 45 (n=4) days. One non-operated animal representing pristine tissue conditions served as the control. The maxillary sinuses with implants were retrieved and further processed to prepare light microscopic ground sections or decalcified sections for immunohistochemical analyses. Results: In paper I seven implants were lost in five patients – six in the maxilla and one in the mandible. All failed implants were Mark III turned implants. The overall implant survival rate was 98.2% with a survival rate of 96.4% for implants with turned surface after a minimum of five months after functional loading. In paper II, one additional oxidized implant failed, giving an overall cumulative survival rate of 94.7 and 99.4%, respectively, after at least five years of functional loading. There was no difference for BoP, PD, or MBL between turned and oxidized implants. A total of two implants, three oxidized and one turned, showed a PD > 3 mm, MBL > 4 mm, and BoP. However, none of these were associated with suppurative infection on examination. In paper III no difference was found between the two implants surfaces used in terms of PD, BoP, MBL, or ABL. Pathological reactions to the sinus membrane were seen in four of the patients (14%). Radiographic signs of sinus pathology were not correlated to either survival rate of the implants or any of the investigated parameters. In the experimental paper IV, bone formation started from the bottom of the sinus floor, sprouting into the granulation tissue along the implant surface under the elevated membrane irrespective of time and surgical technique. Bone formation was not seen in direct conjunction with the sinus membrane. A distinct expression of osteopontin was observed in the serous glands of deeper portion of the lamina propria in direct connection with the elevated sinus membrane and close to the implant within all groups. Conclusion: After more than five years of function in non-grafted patients, oxidized implants had a survival rate higher than turned implants, although this was not statistically significant. No difference was found in MBL, PD, or BoP. Grafting of the maxillary sinus floor with intra- orally harvested bone and delayed placement of either turned or oxidized implants resulted in equally high long-term survival rates, MBL, ABL, and BoP. Pathological findings in the maxillary sinus cavity, in terms of sinus membrane health, are few and not correlated to any of the other investigated parameters. In the experimental study bone formation after sinus membrane elevation with or without additional bone grafts started at the sinus floor and sprouted into the elevated space along the implant surface. Removal of the membrane resulted in less bone formation. The sinus membrane did not seem to present osteoinductive potential in sinus membrane elevation procedures.