Ovarian cancer (OC) has high mortality and chemoresistance rates, necessitating novel prognostic signatures and molecular biomarkers. SUMOylation, crucial in cellular stress responses, alters in various cancers. In this study, using multi-omics data, we characterized the unique features of SUMOylation in OC and revealed the close association between SUMOylation-related genes (SRGs) and OC malignancy. Integrated machine learning identified 22 prognostic SRGs based on the TCGA-OV cohort. Further single-cell analysis refined these findings, pinpointing five SRGs as novel biomarkers closely associated with OC function, metabolism and the tumor microenvironment. In cancer cells, SUMOylation levels regulate the expression of four SRGs (PI3, AUP1, CD200 and GNAS), which are closely associated with the activities of epigenetic regulation process and epithelial mesenchymal signaling. Notably, we discovered that AUP1 overexpression is a risk factor for chemoresistance of OC. In tumor microenvironment, CD8+ cytotoxic T cell with high CCDC80 (another SRG) expression exhibit inhibited cytotoxicity activities. Overall, five SRGs were identified and validated as novel prognostic and therapeutic targets, providing valuable insights for precision medicine of OC.