Objective: To identify independent risk factors for 30-day mortality and construct a prognostic prediction model based on multidimensional clinical indicators in adult patients receiving extracorporeal cardiopulmonary resuscitation (ECPR) who survived at least 72 hours after extracorporeal membrane oxygenation (ECMO) cannulation. Methods: This was a single-center retrospective cohort study. Adult ECPR patients admitted to the Department of Emergency Medicine of the First Affiliated Hospital of Nanjing Medical University between January 2022 and December 2024 were consecutively enrolled. Patients were divided into the survival group (46 cases) and the death group (72 cases) according to their 30-day survival status after ECMO initiation. Baseline characteristics, treatment-related data, and maximum vasoactive-inotropic score within 24 hours and 72 hours after cannulation (VISmax24 h, VISmax72 h) were collected and compared between the two groups. Multivariate logistic regression was performed to establish the prediction model. The area under the receiver operating characteristic curve, calibration curve, and decision-curve analysis (DCA) were used to evaluate discrimination, calibration, and clinical utility of the model, and internal validation was conducted via the Bootstrap method. Results: A total of 118 ECPR patients were included, and the 30-day mortality rate was 61.0% (72/118). Multivariate logistic regression demonstrated that acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score (OR=1.260, 95%CI: 1.064-1.492, P=0.007), coronary heart disease (CHD) (OR=7.464, 95%CI: 1.284-43.386, P=0.025), continuous renal replacement therapy (CRRT) (OR=4.146, 95%CI: 1.400-12.275, P=0.010), and VISmax72 h (OR=1.083, 95%CI: 1.018-1.153, P=0.012) were independent risk factors, whereas delta cardiac troponin-T between day 1 and day 3 after cannulation (ΔcTnT) (OR=0.050, P=0.048) served as an independent protective factor. The prediction model incorporating the above five variables yielded an area under the curve of 0.836 (95%CI: 0.765-0.907). The calibration curve showed high consistency between model-predicted probabilities and actual observed probabilities. Decision curve analysis indicated favorable clinical net benefit of the model. The C-index after Bootstrap internal validation reached 0.891. Conclusion: The prediction model integrating APACHE Ⅱ score, CHD, CRRT, ΔcTnT, and VISmax72 h exhibits satisfactory predictive performance for 30-day mortality risk in adult ECPR patients surviving ≥72 h after ECMO cannulation. Integrating three dimensions including baseline disease severity, organ injury and dynamic hemodynamic recovery, this model provides a quantitative tool for early clinical risk stratification, therapeutic adjustment and doctor-patient communication.