To evaluate presumptive differences in osseointegration at implants installed in healed sites or extraction sockets, supporting either crowns or bridges that were functionally loaded or left unloaded. In six dogs, the mesial roots of the first mandibular molars were treated endodontically. Bilaterally, the teeth were hemisected, and the distal roots extracted. First and second mandibular premolars were extracted as well. After 3 months, the mandibular third and fourth distal roots were extracted after endodontic treatment of the mesial roots. Four implants were installed bilaterally, two in the healed sites corresponding to the second premolar and first molar regions, and two into the extraction sockets. Cobalt-chrome single crowns were prepared and installed in the two most anterior implants, and bridges at the two most posterior implants, bilaterally. A 3-unit bridge was applied to the premolars in the upper jaw only at the loaded sites. All prostheses had a flat occlusal surface and contacts in centric occlusion only at the loaded sites. Three months later, biopsies were retrieved for histological analysis. Higher levels of osseointegration and bone density were observed at the unloaded sites, both at implants installed in healed and post-extraction sites. However, only at implants installed in the post-extraction sites and supporting single crowns, the difference in bone-to-Implant contact was statistically significant. In implant installation immediately following extraction or delayed after three months, osseointegration and bone density were not affected by occlusal contact schemes.