<p>Steroid-induced osteonecrosis of the femoral head (SONFH) is closely associated with coagulation dysfunction, but its molecular mechanisms remain incompletely understood. This study aimed to identify coagulation-related biomarkers for SONFH and evaluate their diagnostic relevance. The GSE123568 dataset from the Gene Expression Omnibus database (GEO) was analyzed, and differentially expressed genes (DEGs) were intersected with coagulation-related genes (CRGs) from the Molecular Signatures Database to identify coagulation-related differentially expressed genes (CRDEGs). Functional enrichment analysis was performed to explore the biological processes and pathways associated with these genes, while protein-protein interaction network analysis and three machine learning algorithms were applied to identify key genes. A nomogram model was constructed and evaluated using bootstrap internal validation. Gene set enrichment analysis, single-sample gene set enrichment analysis, competing endogenous RNA network, and Drug Signature Database analyses were performed to explore potential regulatory mechanisms. The expression levels of the key genes were preliminarily assessed by reverse transcription quantitative polymerase chain reaction (RT-qPCR). In total, 49 CRDEGs were identified and were mainly enriched in immune response, blood coagulation, and complement and coagulation cascades. PECAM1, LGALS3, and ANXA5 were identified as key genes. The prediction model constructed based on these genes showed favorable diagnostic discrimination, with a bootstrap AUC of 0.939 (95% CI: 0.840–0.977). RT-qPCR showed decreased PECAM1 and LGALS3 expression and increased ANXA5 expression in SONFH patients. These findings suggest that PECAM1, LGALS3, and ANXA5 may serve as potential coagulation-related biomarkers for SONFH, providing a basis for further clinical validation and mechanistic investigation.</p>