<p>Gestational diabetes mellitus (GDM) is the most common metabolic disturbance during pregnancy, and currently methods for clinical GDM screening have limitations. This study investigated metabolism-related genes (MRGs) associated with GDM and explored the potential role of these MRGs in the pathogenesis of GDM, for the purpose of providing new reference for the diagnosis and treatment of GDM. Two GDM-related datasets (GSE203346 and GSE103552) were used in this study. MRGs were downloaded from MsigDB. Difference analysis was performed between the GDM and control groups in GSE203346 to obtain differentially expressed genes (DEGs), and then DEGs were intersected with MRGs to obtain differentially expressed metabolism-related genes (DE-MRGs). The expression of DE-MRGs was analyzed using GSE203346 and GSE103552 to obtain characteristic genes. LASSO, SVM-RFE and XGBoost algorithms were used to screen significantly different and consistently expressed genes as hub genes. In addition, the hub genes were further explored by means of correlation analysis, functional analysis, and immune infiltration analysis, and TF-mRNA network and miRNA-mRNA network were constructed. Finally, reverse transcription quantitative polymerase chain reaction (RT-qPCR) was performed to assess the expression of the hub genes. A total of 392 DEGs were obtained, and 18 DE-MRGs were obtained by intersection with MRGs. Then, six genes were obtained as characteristic genes, from which ARG2 and LCMT2 were identified by taking the intersection of results of LASSO, SVM-RFE and XGBoost algorithms. Both ARG2 and LCMT2 exhibited significantly lower expression levels in GDM samples, as observed in public databases and through RT-qPCR. Correlation analysis showed that ARG2 and LCMT2 were positively correlated. The GSEA analysis results showed that the pathways activated by the two hub genes were mainly the ribosome and proteasome pathways. Immune infiltration analysis showed that ARG2 and plasma cells were most significantly negatively correlated. The TF-mRNA network showed that ARG2 and LCMT2 were regulated by YY1. The miRNA-mRNA network showed that ARG2 and LCMT2 were regulated by has-mir-191-5p. We highlighted the potential diagnostic value of two metabolism-related genes associated with GDM, ARG2 and LCMT2, which may be used as biomarkers for GDM diagnosis. Our finding was obtained merely via bioinformatics analysis, it needs to be confirmed by functional studies.</p>