Objective: p53 gene, as “the guardian of the genome”, is the most widely studied tumor suppressor gene. Previous studies have shown that about 50 percent of tumors have P53 dysfunction. This article aims to retrospectively analyze the correlation between p53 rs1625895 polymorphism and the prognosis of patients with diffuse large B-cell lymphoma (DLBCL). Methods: PCR combined with Sanger sequencing were used to detect rs1625895 genotype in 384 DLBCL patients. The relationship between rs1625895 polymorphisms and the clinical characteristics, first-line therapeutic effects and the prognosis of the patients were analyzed. Results: Among all the patients, 2 (0.5%) patients with AA genotype, 34 (8.9%) patients with AG genotype and 348 (90.6%) patients with GG genotype were identified. The patients with different rs1625895 genotypes did not have any difference in terms of age, gender, B symptoms (developing any of the following symptoms: unexplained recurrent fever (often above 38 ℃), night sweats, and unexplained weight loss of 10% within 6 months ), erythrocyte sedimentation rate (ESR), international prognostic index (IPI) and molecular subtype (P>0.05). The overall response rate (ORR) was 82.9% and 82.8% in AA/AG and GG, respectively. There was no significant dif-ference between the first-line therapeutic effects of the two groups (P>0.05). And there was also no difference between A allele carriers and homozygous G allele carriers for the 5-year progression-free survival rate (PFS) (71.8% vs. 62.3%, χ ² = 1.351, P=0.245) and 5-year overall survival rate (OS) (72.2% vs. 64.1%, χ ² = 1.267, P = 0.260). But in the subgroup with Germinal Center B-cell (GCB) type, the patients carrying A allele for rs1625895 had an obviously longer PFS (91.7% vs. 72.7%, χ ² = 4.493,P = 0.034) and OS (91.7% vs. 76.7%, χ ² = 4.246, P = 0.039) compared with the patients homozygous for the G allele. As for the patients with non-GCB subtype, there was no significant difference in PFS and OS between different rs1625895 genotypes (P>0.05). According to whether the first-line regimen contained rituximab or not,the patients were divided into two groups treated with cyclophosphoramide, doxorubicin, vincristine and prednisone (CHOP) or with rituximab and CHOP (R-CHOP). But in both subgroups, there was no significant difference in the 5-year PFS and OS between the AA/AG and GG patients,too (P>0.05). Conclusion: For DLBCL patients receiving CHOP regimen chemotherapy in the first line, p53 rs1625895 cannot predict the clinical efficacy and prognosis of the patients, but in the patients with GCB subtype, this polymorphism may be a prognostic indicator.