Purpose <p>A large number of patients with hip fractures develop postoperative complications. The APACHE II score has been shown its usefulness in surgical patients. We undertook a study to establish the ability of the APACHE score to assess the prediction of postoperative complications in patients over 65 years old with hip fractures, and to compare them with specific scoring systems developed for orthopaedic patients, and the general ASA Score and the Charlson Comorbidity Index (CCI).</p> Methods <p>In order to test the score, we used a prospective, cohort study in a tertiary care hospital with 410 consecutive patients aged ≥65 years old with hip fractures. The ASA score, CCI, the Nottingham Hip Fracture Score (NHFS), the POSSUM-P and the Acute Physiology and Chronic Health Evaluation (APACHE) II scores were tested. All complications up to 6 months post-operatively were registered. We categorized the complications according to the Clavien-Dindo score as minor (1–2) or major (3–5), and we analysed the major post-operative complications.</p> Results <p>The sensitivity and specificity of APACHE score and the P-POSSUM scores were 53.2 and 90.0, and 73.8 and 72.9, respectively. The area beneath the ROC curve was better for P-POSSUM than for APACHE II, but the difference was not statistically significant. Both the P-POSSUM and APACHE II scores were statistically significantly better than ASA, the Nottingham score, and CCI.</p> Conclusion <p>This study showed that APACHE II and POSSUM-P can be used to predict postoperative complications in elderly people with hip fractures.</p>