Abstract In situ normothermic regional perfusion (NRP) and ex situ normothermic machine perfusion (NMP) aim to improve the outcomes of liver transplantation (LT) using controlled donation after circulatory death (cDCD). NRP and NMP have not yet been compared directly. In this international observational study, outcomes of LT performed between 2015 and 2019 for organs procured from cDCD donors subjected to NRP or NMP commenced at the donor center were compared using propensity score matching (PSM). Of the 224 cDCD donations in the NRP cohort that proceeded to asystole, 193 livers were procured, resulting in 157 transplants. In the NMP cohort, perfusion was commenced in all 40 cases and resulted in 34 transplants (use rates: 70% vs. 85% [p = 0.052], respectively). After PSM, 34 NMP liver recipients were matched with 68 NRP liver recipients. The two cohorts were similar for donor functional warm ischemia time (21 min after NRP vs. 20 min after NMP; p = 0.17), UK–Donation After Circulatory Death risk score (5 vs. 5 points; p = 0.38), and laboratory Model for End‐Stage Liver Disease scores (12 vs. 12 points; p = 0.83). The incidence of nonanastomotic biliary strictures (1.5% vs. 2.9%; p > 0.99), early allograft dysfunction (20.6% vs. 8.8%; p = 0.13), and 30‐day graft loss (4.4% vs. 8.8%; p = 0.40) were similar, although peak posttransplant aspartate aminotransferase levels were higher in the NRP cohort (872 vs. 344 IU/L; p < 0.001). NRP livers were more frequently allocated to recipients suffering from hepatocellular carcinoma (HCC; 60.3% vs. 20.6%; p < 0.001). HCC‐censored 2‐year graft and patient survival rates were 91.5% versus 88.2% (p = 0.52) and 97.9% versus 94.1% (p = 0.25) after NRP and NMP, respectively. Both perfusion techniques achieved similar outcomes and appeared to match benchmarks expected for donation after brain death livers. This study may inform the design of a definitive trial.