The aim of the present study was to evaluate healing of revascularised fibula grafts used for mandibular reconstruction using [18F]fluoride ion and positron emission tomography (PET). Sixteen PET studies in 11 fibula grafts were analysed to determine both blood flow and fluoride influx as a measure of vascularisation and osteogenic activity. Two graft failures and three non-unions were encountered and were compared to the successfully healed grafts. In uneventful graft healing, early PET studies (on average 19 days after grafting) showed a significantly increased blood flow to the grafted bone and to the union between the grafts and the mandibles when compared to the reference region of the cervical spine. In contrast, fluoride influx was significantly lower in the grafts when compared to the plating area and the cervical spine. Six months after grafting, blood flow to the grafted bone and the mandibular bone had returned to a level comparable with the reference region. Fluoride influx remained significantly lower in the grafts than in the plating areas or cervical spines. Graft failures were associated with negligible fluoride influx near zero in early PET studies. These results suggest that revascularised fibula grafts provide a low osteogenic potential, presumably due to the pre-existing lack of cancellous bone. The relatively high frequency of non-unions makes meticulous adaptation of the graft and the mandible mandatory, particularly in patients with compromised viability of the recipient bone.