Abstract: Objective: To develop and validate a nomogram for predicting lung injury severity in patients with CTD-ILD by integrating quantitative computed tomography (CT) indicators (SD, NL%, and F%) and clinical characteristics (sex and disease duration), providing a reference for clinical risk stratification. Methods: This retrospective study included 140 patients with CTD-ILD who were divided into mild (Warrick score <8) and moderate-severe (Warrick score ≥8) groups. Quantitative CT images were obtained using a 3D-Slicer. Key parameters were identified using logistic regression analysis and model performance was evaluated using 10-fold cross-validation, bootstrap resampling, and decision curve analysis (DCA). Results: The model identified sex (OR=0.293), SD (OR=1.043), F% (OR=1.708), and NSIP subtype (OR=0.175) as independent predictors. The nomogram showed excellent discrimination (mean AUC of 0.848 in 10-fold cross-validation; AUC of 0.833 in Bootstrap 1000 resamples), good calibration (HL test X2=7.908, P=0.443), and the highest net benefit at a 15% high-risk threshold via DCA, making it suitable for clinical decisions. Conclusion: This study successfully developed a multiparametric prediction model for assessing CTD-ILD severity, demonstrating good discrimination and calibration, thereby offering a valuable reference for individualized treatment and rational resource allocation.