Abstract： Objective： To analyze the application patterns of herbal ironing therapy in ancient texts regarding historical periods， herbal formulas， and disease indications using data mining technology. Methods： Original texts containing "ironing"，"application"，or "hot compress" were extracted from the Acupuncture and Moxibustion Classics Database Platform. Microsoft Excel 2016 was used for frequency statistics，while the database platform and R language were employed for cluster analysis， association rule mining analysis， and complex network analysis. Results： The study included 119 herbal prescriptions involving 107 herbs and 44 diseases from 40 ancient texts. A total of 11 highfrequency herbs（frequency ≥4）were identified，with Allii Fistulosi Bulbus being most prevalent，followed by Natrii Chloridum and Artemisiae Argyi Folium. Herbs were predominantly pungent-warm or bitter-cold in flavor and nature， primarily entering the stomach，lung，and spleen meridians. Therapeutic categories mainly included exterior-releasing， emetic， and interior-warming herbs. High-frequency diseases included ulcerative disorders， syncope / collapse syndromes， and abdominal pain. Cluster analysis revealed six core prescriptions， association rules identified five significant correlations， and complex network analysis demonstrated five strong disease-herb relationships. Conclusion： Herbal ironing therapy represents an external warming method characterized by pungent-warm herbs predominantly entering the stomach meridian， featuring food-grade materials and regionally adapted applications. Treatment primarily targets affected areas and the umbilical/abdominal region，demonstrating efficacy for both cold/heat syndromes and exterior/interior disorders.