Twenty-one consecutive total femoral replacements performed at 2 institutions were reviewed. The surgical indications varied and included bone loss caused by tumor resections (12 patients), failed prior massive partial femoral endoprostheses (6 patients, 5 of whom had prior tumor resections), or difficult to treat fractures and nonunions (3 patients, 2 of whom had prior endoprostheses). Two patients with tumors died within 2 months. The results in the 19 longer-term survivors were satisfactory in 16 patients (good in 7, fair in 9) and poor in 3. Although not commonly used, when this method was indicated it provided an option to hip disarticulation, which often was the only other viable option. Age was the prime determinant of clinical outcome. Average active hip and knee range of motion was better in younger patients (60 years old, 43° and 49°, respectively). All household ambulators required walkers and had a mean age of 81 years; community ambulators requiring 1 cane had a mean age of 46 years; and community ambulators using no walking aids had a mean age of 22 years.