Authors clinically analysed the end results of the 56 elderly patients having fresh femoral neck fracture who were treated with osteosynthesis procedures and 23 endoprosthesis replacement surgery. All these 79 patients were treated between 1970 and 1979 at the St. Marys Hospital, Seoul. The indications of primary endoprosthesis replacement surgery in femoral neck fractured patients were in elderly patients over 65 years, displaced, comminuted sub capital or transcervical fractures, irreducible fractures, pathological fractures and patients having generalized diseases or senile psychosis. The results for the operated hip joints were assessed by duration of hospitalization, early and late post-operative complication. and the Jessels hip grading method at least post-operative one and half year later. The results obtained were as follows: 1. The average age of the patients were 65 years in 56 cases with initial osteosynthetic group, and 73 years in 23 cases of initial endoprosthesis group. 2. The slip down was the most common cause of injuries. The transcervical type of fracture was 54% of all cases. 3. The eligibility of ambulation was average 34 days after surgery in cases having initial osteosynthesis and 18 days in cases having endoprosthesis. 4. Early post-operative complications such as pneumonia and urethritis etc. were more common in cases of endoprosthesis than in osteosynthesis. But we think these complications were not related with surgery itself and transient. 5. Late post-operative complications in cases of osteosynthesis were avascular necrosis of head (16%), nonunion (9%), osteoarthritis (4%), and deep wound infection (2%), but in cases with endoprosthesis, leg length discrepancy (12%), acetabular erosion (4%) and loosening (4%) in order. 6. We performed the second operation in 12 cases among the 56 patients of osteosynthesis because of post-operative complications. 7. Good end results were obtained 90% in 56 osteosynthesis and 91% in 23 endoprosthesis. Conclusively, author's observation time was relatively short (11/2–3 years) and it is impressed that long term study for the prosthetic group is necessary for the more accurate evaluation of their hip function.