Objective To investigate the relationship between the metabolic enzyme genes GSTM3, DPYD, UGT1A1, single nucleotide polymorphisms (SNP) and the clinical progression of colorectal cancer. Methods Totally 690 peripheral blood samples were collected from patients with colorectal cancer from January 2016 to July 2017. TaqMan real-time fluorescence quantitative PCR was used for genotyping of three sites of GSTM3 rs7483, DPYD rs1801159 and UGT1A1 rs3755319. Logistic regression was used to analyze the relationship between three polymorphic sites and colorectal cancer. Results The frequency of GSTM3 rs7483 GG genotype was significantly higher in stage I and II than in stage III and IV (P?0.05). Hierarchical analysis found that in elderly (> 60 years old) patients, the GG carriers in the recessive model were 0.321 times higher than AA + GA genotypes (P?0.05). The poorly differentiated tumors of DPYD rs1801159 AG + GG genotype were 1.398 times of AA genotype (P?< 0.05), especially the poor differentiation of patients with low age≤ 60 years and female patients (P?0.05). No relationship between UGT1A1 rs3755319 and clinical progression of colorectal cancer was found (P?> 0.05). Conclusion Metabolic enzyme genes GSTM3 rs7483 and DPYD rs1801159 may affect the clinical progression of colorectal cancer.