In recent years, bone grafts and bone substitutes have been increasingly utilized underneath barrier membranes to optimize the treatment outcome of bone reconstructive therapy for defects in the alveolar process. In the present study, 4 different filling materials were evaluated in bone defects of similar dimensions in the mandible of miniature pigs. Blood clots and autografts were used as controls. The defects were covered with barrier membranes and allowed to heal for 4, 12 or 24 weeks. Histologic examination demonstrated that bone repair progressed through a programmed sequence of maturation steps closely resembling the pattern of bone development and growth regardless of whether bone grafts or substitutes were present or not. Histomorphometric analysis showed that autologous bone grafts (autografts) had the best osteoconductive properties during the initial healing period, with 39% of newly formed bone inside the membrane-covered defects at 4 weeks of healing. In addition, 87% of the graft surfaces were already covered by bone at this time. Both values were significantly higher for autografts than for the 4 alternative bone fillers (P < or = 0.05). At 12 weeks, these differences were no longer apparent, with all 5 filling materials showing similar values. Among the tested bone substitutes, tricalcium phosphate (TCP) showed a significantly higher percentage of bone fill at 24 weeks of healing. It can be concluded that sites filled with autografts clearly demonstrated the best results underneath barrier membranes in the early phase of healing. As far as degradation and substitution are concerned, TCP showed the most promising results. This filler, however, needs to be tested further in a more demanding animal model. Less favorable results were obtained for coral-derived hydroxyapatite granules and for demineralized freeze-dried bone allografts.