Diffuse Large B-cell lymphoma (DLBCL) is an aggressive disease with heterogeneous outcome and marked variable response to chemotherapy. We assessed promoter hypermethylation (PM) for a panel of tumor suppressor genes in 75 DLBCLs compared to 20 lymphoid hyperplasia (LH) and 30 normal control, using methylation specific PCR. Results were correlated to patients' clinic-pathological characteristics, immunophenotyping, and patients' outcome. DAPK1, RUNX3, MT1G, MGMT, CDH1 and p16 PM were detected in 38.7% (29/75), 49.3% (37/75), 46.7% (35/75), 44% (33/75), 49.3% (37/75) and 42.7% (32/75);respectively, of DLBCL patients compared to LH group (P < 0.05). Aberrant PM of RUNX3, MGMT, CDH1 and p16 was significantly higher in non-germinal central B-cell like (non-GCB) compared to GCB (58.3% vs. 33.3%, 56.2% vs. 22.2, 62.5% vs. 25.9, and 56.2% vs. 18.5%, respectively). PM of studies genes in DLBCL associated significantly with worse survival outcome and resistance to chemotherapy (P ≤ 0.01). In non-GCB group, DAPK1, MT1G, RUNX3, CDH1 and p16 PM associated significantly with reduced DFS (P ≤ 0.004) and OS (P ≤ 0.015). Multivariate analysis indicated that RUNX3 and CDH1 PM were independent prognostic factors for OS (P = 0.03 and 0.04; respectively), while DAPK1, RUNX3 and MT1G PM were independent prognostic factors for DFS (P = 0.002, 0.037& 0.007; respectively). DAPK1, RUNX3, MT1G, CDH1 and p16 PM are promising prognostic and/or predictive markers for non-GCB independent of IPI. Upregulation of those genes using new demethylating agents is a promising approach that sensitize chemoresistant DLBCL patients, especially the non-GCB subtype.