Background <p>Ferroptosis and mitochondrial metabolism are closely associated with the pathological processes of various diseases. However, the role of ferroptosis-related genes (FRGs) and mitochondrial metabolism-related genes (MMRGs) in poor ovarian response (POR) remains unexplored.</p> Methods <p>First, transcriptome sequencing was conducted on ovarian granulosa cell samples from POR patients and controls. Candidate genes were screened through differential expression analysis and consensus clustering analysis. The biomarkers were subsequently screened using machine learning algorithms and receiver operating characteristic (ROC) curves. Nomogram construction and evaluation, enrichment analysis, drug prediction analysis, and molecular docking were subsequently carried out using the biomarkers. Finally, the expression of the biomarkers was authenticated using reverse transcription‒quantitative polymerase chain reaction (RT‒qPCR).</p> Results <p>PLA2G4B and PRKCG were identified as biomarkers by screening. The nomogram demonstrated that these two biomarkers could effectively predict the occurrence of POR. Gene set enrichment analysis (GSEA) revealed that PLA2G4B and PRKCG were both associated with terpenoid backbone biosynthesis. A TF-miRNA-mRNA network was constructed using the biomarkers. For PLA2G4B, dirithromycin had the highest score among the targeted drugs, whereas for PRKCG, bryostatins and enzastaurin had the highest scores. Molecular docking results indicated that their binding energies were less than − 5&#xa0;kcal/mol. RT‒qPCR revealed that PLA2G4B and PRKCG were significantly upregulated in POR samples, which was consistent with the high-throughput sequencing results.</p> Conclusion <p>PLA2G4B and PRKCG have been identified as potential mitochondrial- and ferroptosis-related biomarkers in POR, providing valuable insights for exploring the pathogenesis of POR. These findings may also aid in the development of new diagnostic and therapeutic strategies for POR.</p>