Combined modality treatment of musculoskeletal tumors led to improved patient survival. As survival improves, more consideration is given to the functional outcome of treatment, and interest is focused on the development of less mutilating and extensive surgery. One modality that can reduce patient disability significantly is cryosurgery, as it allows minimally invasive surgery based on marginal resection and tumor interface sterilization instead of wide resection of certain neoplasms. Classical cryosurgery as developed by Marcove involves pouring of liquid nitrogen into the tumor bed. This approach revolutionized the treatment of some tumors such as giant cell tumor of bone, allowing intra-lesional resection to substitute the wide-resection method used up to that time. However, complications of this method of treatment are common, including nitrogen emboli, fractures of the bone due to extensive necrosis and damage to neurovascular elements. A recent development in the field of cryosurgery has been the argon-based system allowing controlled formation of an ice-ball surrounding a metallic probe. The system is computer controlled and allows precise evaluation of the tumor bed interface as well as surrounding structures that need to be protected. Prior to application of this method in humans it is important to ensure that interface sterilization is indeed achieved using cryosurgery. To evaluate this question, a Swarm rat chondrosarcoma was used. Cell viability was assessed following ice-ball formation. Histological evaluation indicated that cell death occurs up to 5 millimeters from the ice-ball if temperatures of -40 degrees Celsius at the metallic probe are achieved. A further evaluation was performed on samples obtained from patients during surgery. A minimum of two freezing cycles was shown to be necessary to achieve tissue viability similar to that of boiled tissue. Twenty-seven patients were operated to date using an argon-based cryosurgery system. The patients included 7 cases of grade I chondrosarcoma, 5 cases of giant cell tumor of bone, 14 cases of a metastatic lytic bone lesion and a single case of osseous-fibrous dysplasia. None of the patients suffered nerve injury during the operation. After a minimal follow-up period of 2 years only two of the surviving patients had a recurrence (a giant cell tumor of the proximal fibula, and the patient with the osseous-fibrous dysplasia whose tumor recurred as a frankly malignant adamantimoma). There were no pathological fractures. This method appears practical and allows close monitoring of the surrounding tissue to reduce the chances of recurrence.