We have evaluated the fate of mineralized osseous implants placed into cranial defects in rats. By 2 weeks, 100% of the defects that had been filled with demineralized bone powder (DBP, 75-250 micrometer) showed bony repair as judged by histomorphometric analysis and incorporation of 45Ca. The DBP was not appreciably resorbed but rather was amalgamated within the new bone. Histomorphometric evaluation of osteo-genesis induced by equal masses of demineralized bone powders of various particle sizes (less than 75, 75-250, 250-450 micrometer) revealed that the smaller particles induced more bone per field than did the larger particles. In contrast, mineralized bone powder (BP) was completely resorbed by 3 weeks, without bony repair of the cranial defect. These specimens contained large multinucleated cells within 7 days and completely resorbed by 3 weeks. It is concluded that (a) demineralized bone powder predictably induces a osteogenic healing of cranial defects, (b) demineralized bone powder is not appreciably resorbed prior to bone induction, (c) the extent of bone induction is a function of the surface area of the demineralized bone implant, and (d) mineralized bone powder undergoes obligatory resorption.