The purpose of this study was to evaluate bone formation and osseointegration using titanium dental implants in the subantral space following surgical implantation of recombinant human bone morphogenetic protein-2 (rhBMP-2). In each of four cynomolgus monkeys, one subantral site was treated with rhBMP-2 (0.19 mg per implant) in an absorbable collagen sponge (ACS). The contralateral site was treated with vehicle in ACS (control). Three months later, two screw-type titanium dental implants were placed into each augmented sinus, and one additional implant was placed immediately anterior to the sinus. Thus, each animal had three experimental sites: rhBMP-2, control, and non sinus. Animals were sacrificed after an additional 3 months, and block sections were harvested and prepared for histometric analysis. Analysis of variance and t tests were used to evaluate differences between experimental conditions. Mean (± SD) vertical bone gain was significantly greater in rhBMP-2 than in control sites (6.0 ± 0.3 versus 2.6 ± 0.3 mm; P .002). Bone density in rhBMP-2 sites averaged 14.4 ± 2.9% versus 13.9 ± 4.6% and 14.1 ± 3.6% for control and nonsinus sites, respectively, without significant differences between experimental conditions. Bone-implant contact in rhBMP-2 sites (41.4 ± 7.7%) was not significantly different from that in control (38.9 ± 12.4%) and nonsinus sites (46.8 ± 10.6%). The present study provides evidence for considerable vertical bone gain in the subantral space following surgical implantation of rhBMP-2, thus allowing placement of dental implants. The newly formed bone appears to be of similar quality and to be as suitable for osseointegration as the residual bone in this nonhuman primate model. Thus, sugrical implantation of rhBMP-2 appears to have clinical utility and may provide a realistic alternative to autogenous bone grafts for subantral augmentation procedures.