Objective <p>Early prediction of the severity of acute pancreatitis (AP) is crucial for clinical decision-making and for improving patient outcomes. In this study, readily available laboratory data and Least Absolute Shrinkage and Selection Operator (LASSO) regression were used to select key variables and construct a nomogram for accurate severity prediction in AP patients.</p> Methods <p>We retrospectively analyzed the clinical data of 965 adult AP patients hospitalized at a hospital between January 2017 and December 2019. LASSO regression was used to identify significant predictors of severe acute pancreatitis (SAP), and a nomogram prediction model was constructed. The performance of the nomogram was evaluated using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, the calibration curve, and the decision curve analysis (DCA).</p> Results <p>Among the 965 patients, the incidence of SAP was 15.96% (n = 154). Eight independent predictors were identified: C-reactive protein, lactate dehydrogenase, hematocrit, monocyte percentage, prothrombin time, <span>D</span>-dimer, glucose, and total cholesterol. These predictive factors were incorporated into a nomogram model, which demonstrated favorable discriminative ability, with AUC values of 0.809 in the training cohort and 0.768 in the validation cohort. The clinical utility of the model was further supported by satisfactory calibration and positive net benefits of DCA.</p> Conclusion <p>The proposed nomogram was highly discriminative in predicting AP severity, indicating its potential clinical value for risk stratification.</p>