One hundred twenty-two patients aged 60–79 undergoing elective total hip or knee arthroplasty were randomized to receive perioperative cognitive training (CT) (n=61) or routine care (RC) (n=61). The primary outcome was the incidence of postoperative delirium (POD). Secondary outcomes included adverse events, postoperative pain within 48 hours, number and duration of POD-positive days, length of stay, and patient satisfaction. Outcomes were assessed up to postoperative day 7 or until discharge. Overall, the study found no significant difference in POD incidence between the CT (9.8%) and RC (6.6%) groups. However, all POD cases in the CT group occurred in those with ≤12 minutes of total training. This suggests that perioperative cognitive training does not significantly reduce POD unless minimum engagement thresholds are met.