Objective To investigate the respective correlations of dihydropyrimidine dehydrogenase (DPD), thymidylate synthase (TS), and thymidine phosphorylase (TP) expression in tumor tissue with the steady-state plasma concentration of 5-fluorouracil (5-FU) and their associations with clinical efficacy in patients with advanced colorectal cancer, so as to provide clues for the "biomarker-pharmacokinetics-clinical efficacy" axis to understand individual differences in 5-FU efficacy. Methods A total of 100 patients with advanced colorectal cancer treated with the standard FOLFOX6 regimen at Ganzhou People's Hospital of Jiangxi Province from January to December 2019 were prospectively enrolled. Immunohistochemistry was used to detect the expression (positive/negative) of DPD, TS, and TP in tumor tissues. The steady-state plasma concentration was measured at the 24th hour of continuous 5-FU infusion in the first cycle. After 6 cycles of chemotherapy, the objective response rate was evaluated according to the Response Evaluation Criteria in Solid Tumors. The objective response rates were compared between patients with different expression statuses, and point-biserial correlation analysis was used to examine the association between expression (binary variable) and plasma concentration (continuous variable). Results The objective response rates in patients with positive expression of DPD, TS, and TP proteins were 32.43%, 37.21%, and 35.56%, respectively, all lower than those in patients with corresponding negative expression (63.49%, 61.40%, 65.45%, respectively, all P<0.05). Point-biserial correlation analysis showed that expression of DPD, TS, and TP was negatively correlated with steady-state plasma concentration of 5-FU (rpb=-0.786,P<0.001; rpb=-0.752, P=0.002; rpb=-0.737, P=0.010). Conclusion Positive expression of DPD, TS, and TP in tumor tissue is associated with lower steady-state plasma concentration of 5-FU and poorer clinical efficacy.This suggests that high expression of these proteins may collectively contribute to inferior treatment outcomes through distinct mechanisms, indicating their potential as negative predictive biomarkers.