The aim of the study was to assess quantitatively the amount of new bone formed in the early stages of healing of intramembranous and endochondral autogenous bone grafts so as to gain further insight into their integration with host bone. Eighteen critical size defects were created in the parietal bone of nine New Zealand White rabbits. In the experimental group (five rabbits), each rabbit was grafted with intramembranous bone in one defect and with endochondral bone in the other. In the control group (four rabbits), one defect was left empty (passive control) and the other was grafted with rabbit skin collagen (active control). After 14 days, the rabbits were killed and the defects were prepared for histological analysis. Serial sections were made across the whole defect. Each defect was divided into five regions spaced 1500 microns apart. Two sections were randomly drawn from each region. Quantitative analysis was performed on 100 sections using an image analyser computer software system to assess the amount of new bone formed in each defect. No bone was detected across the defect in either the active or passive controls. One-hundred-and-sixty-six per cent more new bone was formed in defects grafted with intramembranous bone than those grafted with endochondral bone. This represented an extremely significant difference (P < 0.0001, unpaired t-test) between the two groups. The results show that intramembranous autogenous bone produced more bone than the endochondral bone when grafted in the skull. Clinically, it is recommended that intramembranous bone is used to replace lost membranous bone in the oral cavity, as well as in skull defects, whenever possible.