Background: The aging of the Chinese population is expected to lead to an increase in nonagenarians and centenarians. The mortality rate in nonagenarian hip fracture patients is equivalent to the mortality rate in the average population at 5 years after injury. It is imperative to evaluate 5-year mortality in this small but very challenging subgroup of patients to optimize patient management. The primary purpose of the current retrospective study was to compare five-year survival in patients aged over 90 years who received arthroplasty or nonoperative treatment for femoral neck fracture during a 16-year period. Methods: From January 1998 to December 2014, all consecutive nonagenarian and centenarian patients with femoral neck fracture admitted to our hospital were included in the evaluation. The primary outcome was defined as thirty-day, 1-year, 3-year, and 5-year mortality after injury. Survival analysis was performed with the Kaplan-Meier method. Using the log-rank test, stratified analyses were performed to compare differences in the overall cumulative mortality and mortality at three time points (1 year, 3 years, and 5 years) after injury and differences in survival distributions. Results: Over the 16-year study period, the arthroplasty group and the nonoperative treatment group included 33 and 53 patients, respectively. The long-term survival probability of the arthroplasty group was significantly higher than that of the nonoperative treatment group (p = 0.002). The survival time of the arthroplasty group was significantly higher than that of the nonoperative treatment group (median (P 75-P 25) = 53 (59) versus median (P 75-P 25) = 22 (52), p = 0.001). The mortality differences, except for 30-day mortality, at five time points (1, 2, 3, 4, and 5 years) between the nonoperative group and arthroplasty group were significant. The stratified analyses of overall cumulative mortality and mortality at three time points (1, 3, and 5 years) after injury demonstrated that the nonoperative treatment group had significantly higher cumulative mortality than the arthroplasty group. Conclusions: Our study demonstrates that arthroplasty is more likely to improve long-term survival in femoral neck fracture patients aged over 90 years than nonoperative treatment. It can be expected that nearly half of patients will survive more than 5 years after surgery.