PURPOSE: People with obesity often develop non-alcoholic fatty liver disease (NAFLD) and are at high risk of progression to non-alcoholic steatohepatitis (NASH). Few therapies are effective other than bariatric surgery. We therefore analyzed data from duodenal-jejunal bypass liner (DJBL) patients regarding steatosis, fibrosis, and NASH. METHODS: Consecutive DJBL patients with type 2 diabetes underwent standardized assessments up to device removal at 48 weeks. These included aspartate and alanine transaminase (AST, ALT), controlled attenuation parameter (CAP, for steatosis), and liver stiffness measurement (LSM, for fibrosis). The NAFLD fibrosis score (NFS), fibrosis-4 score (FIB4), and enhanced liver fibrosis (ELF) test were also used to assess fibrosis and the Fibroscan-AST (FAST) score to assess NASH. Mixed models were used and missing data were accounted for with multiple imputation. RESULTS: Thirty-two patients (18 female, mean age 55.1, mean BMI 40.2 kg/m2) were included. After 48 weeks, the change compared to baseline with 95% CI was a factor 0.74 (0.65 to 0.84) for AST, 0.63 (0.53 to 0.75) for ALT, and a difference of - 0.21 (- 0.28 to - 0.13) for FAST, all with p < 0.001. Fibrosis based on LSM, NFS, and ELF did not change whereas FIB4 exhibited slight improvement. Eight DJBL were explanted early due to device-related complications and eight complications led to hospitalization. CONCLUSIONS: One year of DJBL therapy is associated with relevant improvements in non-invasive markers of steatosis and NASH, but not fibrosis, and is accompanied by a substantial number of complications. Given the lack of alternatives, DJBL deserves further attention.