Objective:To identify independent risk factors for chronic critical illness (CCI) secondary to severe pneumonia and to develop and validate a clinical prediction model.Methods:A retrospective cohort study was conducted using electronic medical records from 415 patients with severe pneumonia admitted between January 2023 and March 2024. Patients were randomly divided into a training set ( n = 290) and a validation set ( n = 125) at a 7:3 ratio. Univariate and multivariate logistic regression analyses were used to identify independent risk factors, and a nomogram was constructed. The model’s discriminative ability, calibration, and clinical utility were assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). Results:The overall incidence of CCI was 23.13% (96/415). Multivariate analysis identified five independent predictors: virus infection ( OR = 13.00, 95% CI: 5.07–33.35, P < 0.001), mechanical ventilation ≥72 hours ( OR = 8.06, 95% CI: 3.68–20.09, P < 0.001), neutrophil-to-albumin ratio (NAR) ( OR = 27848, 95% CI: 193.93–5542274.11, P < 0.001), oxygenation index ( OR =1.09, 95% CI: 1.01–1.09, P < 0.001), and age ( OR = 0.94, 95% CI: 0.91–0.97, P < 0.001). The model demonstrated excellent performance in both sets: training set AUC = 0.96 (95% CI: 0.94–0.98), sensitivity 0.93, specificity 0.89, Brier score 0.09; validation set AUC = 0.93 (95% CI: 0.88–0.98), sensitivity 0.89, specificity 0.64, Brier score 0.13. Calibration curves showed high consistency between predicted and observed risks (mean absolute error < 3%), and DCA indicated significant net clinical benefit within the threshold probability range of 15–60%. Conclusions:The developed prediction model integrates etiological, inflammatory, metabolic, and respiratory support parameters and demonstrates outstanding predictive performance (AUC > 0.90). It may serve as a quantitative tool for early risk stratification and intervention in patients with severe pneumonia. Further multicenter external validation and exploration of integrating dynamic biomarker monitoring are recommended.