Prolonged high-concentration cortisol exposure may impair cognitive function, but its mechanisms and risk factors remain unclear in humans.Using Cushing's disease patients as a model, this study explores these effects and develops a predictive model to aid in managing high-risk patients.This single-center retrospective study included 107 Cushing's disease patients (January 2020-January 2024) at the First Medical Center of the PLA General Hospital. Cognitive function, assessed using the Montreal Cognitive Assessment, revealed 58 patients with cognitive impairment and 49 with normal cognitive function. Patients were divided into training (n = 53) and validation cohorts (n = 54) for constructing and validating the predictive model. Risk factors were identified via univariate analysis and least absolute shrinkage and selection operator regression, and a nomogram prediction model was developed. Performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).Cortisol AM/PM ratio, 8 a.m. cortisol concentration, body mass index, and fasting plasma glucose were significant risk factors for cognitive impairment. The nomogram demonstrated strong predictive ability, with ROC values of 0.80 (training) and 0.91 (validation). DCA indicated superior clinical utility compared to treating all or no patients.This study confirms the significant impact of prolonged high cortisol exposure on cognitive function and identifies key risk factors. The nomogram model offers robust performance, providing a valuable tool for managing Cushing's disease patients' cognitive health and informing strategies for other cortisol-related disorders.© The Author(s) 2025.