OBJECTIVE We intend to evaluate the efficacy of salvage treatments for relapsed or refractory diffuse large B-cell lymphoma (R/R DLBCL) through meta-analysis. R/R DLBCL trials were divided into two groups based on eligibility for autologous stem-cell transplantation (ASCT), and meta-analysis of each group was performed. Random effects models were used to estimate the 1-year progression-free survival (PFS) rate, and chimeric antigen receptor (CAR) T-cell therapy was used as reference treatment. RESULTS Twenty-six ASCT-eligible cohorts from 17 studies comprising 2,924 patients and 59 ASCT-ineligible cohorts from 53 studies comprising 3,617 patients were included in the pooled analysis. In the ASCT-eligible group, the pooled 1-year PFS rate was 0.40 (95% confidence interval [CI], 0.15 to 0.65) for the CAR T-cell group and 0.34 (95% CI, 0.30 to 0.37) for the group with chemotherapy followed by ASCT intention. The two treatments were not significantly different in meta-regression analysis. In the ASCT-ineligible group, the pooled 1-year PFS was 0.40 (95% CI, 0.35 to 0.46) for CAR T-cell, and the highest primary outcome was 0.47 (95% CI, 0.37 to 0.57) for the tafasitamab group. CAR T-cell therapy showed significantly better outcomes than chemotherapy and therapies based on ibrutinib, lenalidomide, and selinexor. However, loncastuximab, polatuzumab plus bendamustine and rituximab, and the tafasitamab group showed no different efficacy than CAR T-cell therapy after adjusting for median number of previous lines of treatment. CONCLUSIONS Although several regimens were crudely grouped for classification, CAR T-cell therapy did not outperform chemotherapy followed by ASCT in the second-line setting or several recently developed agents in the ASCT-ineligible setting.