ABSTRACT Alopecia areata (AA) is an immune‐mediated chronic alopecia disease, but its specific pathogenesis is unclear. Gene expression data for AA patients (AAs) and healthy controls (HCs) were retrieved from the GEO database, and the differentially expressed genes (DEGs) between AAs and HCs were identified. Then, GO, KEGG and GSEA analysis were performed. A PPI network for the DEGs was then constructed to screen for hub genes, which were validated by three additional datasets. Subsequently, the potential miRNAs interacting with the hub genes were obtained through TarBase and miRNet. The differentially expressed lncRNAs (DElncRs) were obtained for subcellular localisation analysis, and the DElncRs located in the cytoplasm were further screened to identify miRNAs that interact with them. The shared miRNAs interacting with the hub genes and lncRNAs were used to construct a network of mRNA‐miRNA‐lncRNA interactions. Lastly, ROC analysis was performed to evaluate the potential diagnostic value of the hub genes and DElncRs identified. A total of 173 DEGs were obtained, mainly enriched in cytokines, chemokines, hair follicle development and hair cycle related signalling pathways. Through PPI screening and validation based on 3 additional datasets, 24 hub genes were finally yielded. Of them, five hub genes were upregulated and the potential miRNAs that interact with these five hub genes were identified. Additionally, 26 DElncRs were obtained, including 9 upregulated lncRNAs located in the cytoplasm that were predicted to interact with the miRNAs. Finally, an mRNA‐miRNA‐lncRNA regulatory network was constructed using five hub genes, four lncRNAs and their shared five miRNAs. The regulatory relationship between CD8A, mir‐185‐5p and FOXD2‐AS1 might be crucial in AA pathogenesis, with CD8A and FOXD2‐AS1 exhibiting diagnostic potential. CD8A and FOXD2‐AS1 may serve as potential therapeutic targets in AA.