Background <p>Cervical cancer (CC) is a common malignant tumor of the female reproductive system. Currently, patients with advanced or recurrent CC have poor prognosis. The aim of this investigation is to pinpoint key genes that might influence the prognosis of CC patients.</p> Methods <p>CC-related datasets were obtained from TCGA and GEO databases. We employed differential expression analysis and univariate Cox regression to identify the genes associated with CC prognosis. The prognostic value of hub gene MFSD3 was then evaluated via Kaplan–Meier survival curve and multivariate Cox regression analyses. Functional analysis and immune cell infiltration analysis were then conducted. Finally, the promising targeted drug was predicted and analyzed through molecular docking.</p> Results <p>Compared to normal cervical samples, MFSD3 was significantly upregulated in CC samples. The elevated levels of MFSD3 were linked to a poor prognosis and was an independent prognostic biomarker for CC. Functional enrichment analysis showed that MFSD3 was involved in certain energy metabolism related signaling pathways. Immune cell infiltration results showed that high MFSD3 expression was associated with infiltration of various immune cells, such as M2 macrophages. The results of molecular docking, molecular dynamics simulation and cellular thermal shift assays showed that two anticancer drugs (Palbociclib and Dabrafenib) can stably bind to the MFSD3 protein.</p> Conclusion <p>Our study revealed that MFSD3 may serve as a potential prognostic biomarker for CC.</p>