Implant surface morphology regulates osteoblast phenotypic expression. Osteoblast sensitivity to non-biologic surfaces suggests that native bone surface features may also affect osteoblast response. To test this, MG63 osteoblast-like cells were grown for 7 days on bovine cortical bone wafers pretreated with rat bone marrow osteoclasts for 0, 10 or 20 days. Response to osteoclast-treated surfaces was compared to the response of MG63 cells to titanium surfaces with smooth and rough microtopographies. Cell number, differentiation (alkaline phosphatase activity and osteocalcin levels), and local factors (PGE? and TGF-fi 1) were measured in confluent cultures. Compared to culture on plastic, cell number was reduced on all three types of bone wafers; this eflect was dosedependent with increasing resorption of the surface. Alkaline phosphatase specific activity was increased ( P < 0.05) on all three surfaces compared with plastic, but this increase was not dependent on resorption time, indicating this parameter was sensitive to the surface (bovine bone vs. plastic) but not to osteoclast-resorption. There was a direct correlation between the area of the bone surface resorbed and the amount of osteocalcin, TGF-PI and PGEz (R' = 0. 8025, 0.8689, 0.8896, respectively). With 20 days of osteoclast pretreatment, there was a 20-fold increase in osteocalcin over plastic and a 7-fold increase over cultures on untreated bone wafers. Similar increases were found for TGF-PI and PGEz. Thus, surface changes resulting from osteoclast pretreatment have a strong effect on osteoblast phenotypic expression, and suggest that microtopography may play a role.