Objective To construct and preliminarily validate a CT-based patient-specific finite element (FE) model for locating skeletal fragile sites and estimating safe impact thresholds in elderly patients with a second fracture, so as to provide a quantitative reference for individualized fracture prevention assessment by nurses. Methods Ten patients with a second fracture (7 hip and 3 distal radius fractures) admitted to Chinese PLA General Hospital between January 2024 and March 2026 were retrospectively enrolled. Three-dimensional FE models were reconstructed from contralateral CT images obtained at the first fracture to simulate the mechanical process of low-energy injury-induced fracture. Using the affected-side CT images obtained at the second fracture as the reference, agreement between model prediction and actual fracture was evaluated in three dimensions: number of fracture fragments, fracture classification, and key angles (Pauwels angle for the hip and ulnar deviation angle for the radius). High-risk stress concentration areas and fracture thresholds were also generated. This was designed as a feasibility study, and all statistical results were regarded as hypothesis-generating rather than confirmatory. Results The overall agreement for the number of fracture fragments was 9/10 (hip 7/7, distal radius 2/3), and that for fracture classification was 9/10. The mean absolute error of the Pauwels angle was (4.4±2.9)°, and the predicted values were positively correlated with the measured values (r=0.978, n=7); this correlation was exploratory. The mean absolute error of the ulnar deviation angle was (6.4±4.4)°. The model identified the most fragile site and the fracture threshold for each patient, and all fracture thresholds were lower than the corresponding estimated fall impact forces. Conclusions In this small sample, the FE model based on contralateral CT images showed preliminary feasibility in locating skeletal fragile sites and quantifying safe impact thresholds in elderly patients with a second fracture, which may provide a quantitative reference for individualized fracture prevention assessment by nurses. Given the limited sample size and retrospective design, these findings need to be confirmed in larger and experimental studies.