Purpose: To assess the course of the stability and the failure rate of dental implants placed in the partially edentulous maxillae of minipigs. Materials and Methods: Three months after tooth removal, implants were placed in 9 minipigs. Six implants (XiVE; Friadent, Mannheim, Germany) were placed on each side of the posterior maxilla after preparation of the implant sites either by an osteotome technique or with spiral drills. Implant stability was assessed by resonance frequency analysis (RFA) at the time of placement, at second-stage surgery (which took place after a healing periods of 1, 2, 3, 4, or 5 months), and after a loading period of 6 months. Results: Implant stability was significantly influenced by the healing period (P = .007). Implant stability decreased after 1 to 3 months of healing for both of the placement techniques and increased after a healing period of 4 months. After implant site preparation by an osteotome technique, 6 of 12 immediately loaded implants, 18 of 24 implants loaded after healing periods of 1 to 3 months, and 1 of 18 implants loaded after a healing period of 4 or 5 months were lost. After implant site preparation using spiral drills, 7 of 12 immediately loaded implants, 12 of 24 implants loaded after healing periods of 1 to 3 months, and 2 of 18 implants loaded after healing periods of 4 or 5 months were lost. Broad overlapping of confidence intervals for the number of implant failures revealed that there was no relevant difference between immediate and early functional loading for either of the 2 techniques. Discussion and Conclusion: Implant loading after healing periods of 1 to 3 months did not improve implant survival compared to immediate loading in the posterior maxillae of minipigs. Not until a healing period of 4 months was reached did implant stability begin to increase. Only when functional loading was started at this point in time was maximal implant survival achieved.