One hundred fourty four older patients with hip fracture scheduled for unilateral hip arthroplasty or closed reduction and intramedullary nailing under spinal anesthesia were randomized to receive intranasal saline (Control, n=48), intranasal insulin 20U (I-20, n=48), or intranasal insulin 40U (I-40, n=48) given three times (19:00 the day before surgery, 50 minutes before anesthesia, and 19:00 on the day of surgery). The primary outcome was the incidence of postoperative delirium (POD) within postoperative days 1–3, assessed daily with the CAM-ICU; secondary outcomes included day-specific POD incidence, postoperative pain (VAS), cerebrospinal fluid (CSF) glucose and lactate levels, fingertip blood glucose at predefined perioperative time points, and hypoglycemic events. Outcomes were assessed up to 3 days after surgery. Overall, the results of the study revealed that intranasal insulin at both 20U and 40U significantly reduced POD incidence versus placebo without lowering peripheral blood glucose, while 40U modestly increased CSF glucose compared with control. Taken together, these findings suggest that low-dose (20U) intranasal insulin is a simple, non-invasive and potentially optimal regimen to prevent early postoperative delirium in older hip fracture patients without causing hypoglycemia.