Background <p>Glioma was a kind of malignant tumor associated with high mortality and recurrence. Therefore, it was urgent to establish effective prognostic models to guide clinical treatment in glioma. Protein lactylation was discovered in various malignant tumor, but only a few studies on glioma focused on protein lactylation.</p> Methods <p>The expression levels of lactylation-related genes were identified from the TCGA database. A lactylation-related prognostic signature was established using various combinations of 10 different excellent machine learning methods. The prognostic value of the signature was assessed and further validated in the CGGA cohorts. Patients were divided into two groups according to the risk score (RS). Independent prognostic value assessment, pathways enrichment analysis and protein-protein interaction analysis were conducted. Finally, we verified the functions of C19orf53 through vitro experiments.</p> Results <p>A robust lactylation-related prognostic signature of low-grade glioma (LGG) was established, which was consisted of 14 genes. Patients with higher RS had poorer clinical outcomes in all the cohorts. More immune-related and pro-cancer pathways were enriched in high-RS subgroup. Moreover, the 14 lactylation-related prognostic genes had close interaction relationships, and 11 of them had independent prognostic value. Vitro experiments proved that shRNA-mediated C19orf53 down-regulation impeded the migration and proliferation of LGG cells.</p> Conclusions <p>The lactylation-related prognostic signature exhibited robust predictive efficiency in LGG, providing a new perspective for the prognosis evaluation of LGG patients and the subsequent studies on therapeutic targets.</p>