To investigate the effect of implant type after direct contamination, a hydroxyapatite-coated or noncoated Ti-6A1-4V implant was inserted into both tibiae of 32 New Zealand White rabbits. Prior to implantation, the left tibia was contaminated with increasing concentrations of Staphylococcus aureus (10(2)-10(5) colony-forming units), ranging from very low (10(2)) to relatively high (10(5)). Four weeks after implantation, half of the tibial bone adjacent to the implant was harvested for bacteriological examination. Bacterial counts were quantified by plating serial dilutions. For the histological evaluation, sections of the implant with the remaining tibia were examined by semiquantitative scoring of infection parameters. The bacteriological data showed the inoculum dose and implant type to have a significant effect on the culture outcome: more bacteria were retrieved from the hydroxyapatite-coated implants than from the noncoated titanium implants. Histological evaluation showed an increased score for the infected left tibiae compared with their contralateral control. In addition, with increasing inoculum dosage, the difference between the two types of implant increased. We demonstrated that infections can occur with biocompatible, noncemented implants and that they are related to the dose of the original inoculum. Bacteria were more likely to grow onto or next to the hydroxyapatite implants than on titanium implants and resulted in a more severe histopathological characterization of infection.