Introduction. Aseptic loosening is the most common mode of failure of massive endoprostheses. Introduction of Hydroxyapatite coated collars have reduced the incidence of aseptic loosening. However bone growth is not always seen on these collars. Objectives. The aims of our study were to determine the extent of osseous integration of Hydroxyapatite coated collars, attempt a grading system for bone growth and to determine the effect of diagnosis, surgical technique and adjuvant therapy on bone growth. Methods. We reviewed the records and radiographs of 58 patients who had a massive endoprosthesis implanted by two surgeons in our unit over the last five years. Revision surgeries were recorded separately. Bone growth was graded 1–4 based on appearance in antero-posterior and lateral radiographs. Results. Three groups were identified. Group 1-Resections for primary bone tumours (33 patients), Group 2-resections for metastatic bone disease (22 patients) and Group 3- Resections for non tumour indications (3 patients). Overall, 60% of patients had grade 1, 12% had grade 2, 19% had grade 3 and 9% had grade 4 osteointegration. Grade 3 or 4 Collar osteointegration was found in 37% of patients in Group 1, 9% in group 2 and 67% in group 3. 5% of patients with grade 1 integration, 100% patients with grade 2 integration and none of the patients with grade 3 or 4 integration underwent revision for aseptic loosening. Appearance or widening of a gap between the resected bone end and the collar indicated loosening and impending revision. Proximal humeral replacements had the lowest rate of osteointegration (12%). Adjuvant therapy did not affect osteointegration. Conclusion. Osteointegration of collars is seen more often after resection of primary bone tumours. The role of collars in metastatic tumour surgery is questionable. Our radiographic grading system of bone growth predicted aseptic loosening