It is often assumed that there is a direct relationship between the bone density adjacent to an implant, as revealed by radiographs, and the percent histologic osseointegration. Moreover, the lack of standardized methods for evaluation of histologic preparations makes it difficult to compare published studies, especially as little is known about the variables that influence these measurements. In this animal study, computer-assisted lineal analysis was used to evaluate the effects of subject, tooth position, and implant surface site on measured bone density and osseointegration in a bone augmentation experiment. Three sites--coronal lingual, apical lingual, and apical facial--were analyzed around each of 6 (3.75 x 8 mm) threaded machined titanium implants, as well as the apical facial site of 21 other implants placed in the mandibular premolar area of 5 dogs. In all sites, a progressive decrease in bone density was observed from bone adjacent to the implant to that at the titanium implant surface. There was an animal effect on osseointegration, but there were no differences between the mandibular premolar locations (second, third, and fourth). Most importantly, there were significant measurable effects attributable to the surface site examined. The need for carefully standardized histologic evaluations is established. Keywords: bone density, endosseous dental implants, histology, histomorphometry, osseointegration