PURPOSE: The aim of the present study was to evaluate whether there was a difference between machined and TiO(2)-blasted implants regarding survival rate and marginal bone loss during a 5-year observation period. MATERIALS AND METHODS: A total of 133 implants (Astra Tech Dental Implants; Astra Tech AB, Mölndal, Sweden) were placed in 50 patients at 6 centers in 4 Scandinavian countries. Forty-eight implants were installed in the maxilla and 85 implants in the mandible. A randomization and a stratification were done, so that each fixed partial prosthesis was supported by at least 1 machined and 1 TiO(2)-blasted implant. The implant-supported fixed partial prostheses (ISFPP) were fabricated within 2 months after postoperative healing. A total of 52 ISFPP (17 maxillary, 35 mandibular) were inserted. The patients were clinically examined once a year for 5 years. At the annual follow-up, biological as well as technical complications were recorded. RESULTS: Of the 133 implants placed, 3 were reported as failed after 5 years of follow-up, resulting in an overall cumulative survival rate of 97.6%. The cumulative implant survival rates were 100% for the TiO(2)-blasted implants and 95.1% for the machined implants. No significant difference in survival was, however, found between the machined and TiO(2)-blasted implants after 5 years. The mean marginal bone loss in the maxilla was 0.21 +/- 0.83 mm (SD) for the machined implants and 0.51 +/- 1.11 mm (SD) for the TiO(2)-blasted implants during the 5-year observation period. In the mandible, the mean marginal loss was 0.22 +/- 1.13 mm for the machined implants and 0.52 +/- 1.07 mm for the TiO(2)-blasted implants from baseline to the 5-year examination. No significant difference in marginal bone loss between the 2 surface groups was found during the 5-year observation period. CONCLUSIONS: The present study shows good 5-year results with small ISFPP in the mandible, as well as in the maxilla. No significant differences were found in failure rate and marginal bone loss around implants with a machined rather than a TiO(2)-blasted surface. J Prosthodont 2001;10:2-7.