Purpose: The present study was undertaken to determine the feasibility of using primary stability as a predictor of implant success in patients whose implants were immediately loaded. Materials and Methods: The study included 40 patients, in whom a total of 190 implants were placed, 102 in maxillary sites and 88 in mandibular sites. All were loaded within 72 hours of placement. Sixteen patients were completely edentulous in the mandible and/or the maxilla. The remaining 24, who were partially edentulous, received fixed partial dentures or single-implant restorations. All of the definitive implant restorations were screw retained. The criterion for loading was clinical judgment of primary stability, verified by a "screw test." Impressions were made after implant placement to facilitate the fabrication of a laboratory-made heat-processed provisional restoration from acrylic resin. Following a 4-month period for osseointegration and soft tissue healing, definitive fixed prostheses were fabricated. Results: There were no surgical complications. After 1 to 2 years, all 190 implants had survived and were considered 100% successful, as determined by independent testing of mobility and radiographic evidence of osseointegration. In 4 patients, fracture of the provisional restoration occurred during the healing period. Discussion: Clinical research has shown that immediate loading is a viable treatment modality. The favorable success rate reported in this study for rough-surfaced implants suggests that adherence to a protocol, an important parameter of which is primary stability above 32 Ncm, can lead to osseointegration. Conclusion: The results of this limited investigation suggest that patients who are partially or completely edentulous may be immediately restored with implants and fixed provisional restorations, provided that the dental implants are adequately stable immediately after their surgical placement. This alternative therapeutic approach did not appear to affect the up-to-2-year survival of the implants in this patient population.