Serous ovarian cancer (SOC) accounts for >50% of all epithelial ovarian cancers. However, patients with SOC present with various degrees of response to platinum‑based chemotherapy and, thus, their survival may differ. The present study aimed to identify the candidate genes involved in the carcinogenesis and drug resistance of SOC by analyzing the microarray datasets GDS1381 and GDS3592. GDS1381 and GDS3592 were downloaded from the Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/gds/). A total of 219 differentially expressed genes (DEGs) were identified. Potential genes that may predict the response to carboplatin and, thus, the prognosis of SOC were analyzed. The enriched functions and pathways of DEGs included extracellular region, extracellular space and extracellular exosome, among others. Upon screening the upregulated and downregulated genes on the connectivity map, 10 small‑molecule drugs were identified that may be helpful in improving drug sensitivity in patients with ovarian cancer. A total of 30 hub genes were screened for further analysis after constructing the protein‑to‑protein interaction network. Through survival analysis, comparison of genes across numerous analyses, and immunohistochemistry, GNAI1, non‑structural maintenance of chromosomes (non‑SMC) condensin I complex subunit H (NCAPH), matrix metallopeptidase 9 (MMP9), aurora kinase A (AURKA) and enhancer of zeste 2 polycomb repressive complex 2 subunit (EZH2) were identified as the key molecules that may be involved in the carcinogenesis and carboplatin resistance of SOC. In conclusion, GNAI1, NCAPH, MMP9, AURKA and EZH2 should be examined in further studies for the possibility of their participation in the carcinogenesis and carboplatin response of SOC.