This study investigated the concept of using plates to attach endoprostheses to bone after segmental resection for bone tumours in an animal model. Titanium alloy plates integrated with the prosthesis and coated with hydroxyapatite were attached to bone by screws. This type of uncemented fixation relied on the induction of periosteal bone formation into and around the plates to secure the implant to bone. Two, three, and six-slotted plate designs were investigated. On retrieval, each plate was securely fixed by new bone. Bone apposition on the external surface of the plates occurred through a combination of periosteal bone production, invasion of bone through slots in the plate, and bone growth over the ends of the plates. Most plates became incorporated into a remodelled cortex. Higher bone turnover rates (microm day(-1)) were seen in bone in the slots of the plate compared with normal cortical bone turnover (p < 0.05). Significantly higher rates of turnover were measured beneath slotted parts of the plates compared with regions below the unslotted parts (p < 0.05). The cross-sectional area of bone surrounding the six-plate implant design was significantly higher than that of the three-plate (p < 0.05) and two-plate (p < 0.05) designs. In addition, significantly more bone formed adjacent to the six-plated implant design compared with that in the contralateral limb (p = 0.002). However, no significant difference was found when the total cortical area around the three-plated design was compared with that of the contralateral limb (p = 0.63). In contrast, significantly less bone was measured adjacent to the two-plate design than in the untreated limb (p = 0.001). Image analysis also demonstrated increased cortical porosity adjacent to the six-plate design compared with the three-plate (p = 0.004) and two-plate (p < 0.05) designs. Finite element analysis demonstrated that the six and three-plate designs increased the second moment of area compared with that in the left tibia (p = 0.003 and 0.066, respectively). However, the attachment of the more flexible two-plate design did not significantly increase the second moment of area compared with that in the contralateral limb (p = 0.235). It was concluded that due to both mechanical and biological effects, the hydroxyapatite-coated plate designs generated new bone that enhanced fixation and encouraged plate integration into the load-bearing structure of the cortex. This method of fixation may be an alternative to the use of intramedullary cemented stems in patients requiring bone tumour implants and may be the only way to preserve the joint in difficult cases where only short segments of bone remain.