Background <p>Colorectal cancer (CRC) is a common malignant tumor worldwide. The cholesterol synthesis (CS) pathway is crucial in the occurrence and development of cancer. This study aims to predict the prognosis of CRC patients based on the cholesterol synthesis-related genes (CSRGs).</p> Methods <p>The patient data of CRC were downloaded from the TCGA and GEO databases, and the CSRGs were downloaded from Genecards. In the TCGA-CRC training set, univariate Cox regression analysis was conducted on the CSRGs, and subtype classification was performed through consensus clustering. Combined with the PPI network and regression analysis, key CSRGs were identified to establish a prognostic model. ROC curves and Kaplan–Meier survival analysis were used to evaluate the model and validate it in the GSE17538 validation set. At the same time, immune analysis and drug sensitivity analysis were conducted. Finally, the functions of these characteristic genes were investigated in an in vitro cell model.</p> Results <p>The TCGA-CRC was divided into two subtypes. A 10-gene Cholesterol Synthesis-related Risk Signature (CSRS) was constructed. The patients were grouped according to the median value of the CSRS. The high-CSRS group had a poorer prognosis, and the abundance of macrophages, neutrophils, and TIL was higher in this group. The drug sensitivity prediction indicated that several candidate drugs (such as Linsitinib) might affect the progression of CRC through unique mechanisms. In vitro experiments demonstrated that EEF1A2 could promote the malignant progression of tumors.</p> Conclusion <p>The results of this project provide some guidance for elucidating potential CS-related biomarkers for predicting prognosis in CRC patients.</p>