Non-alcoholic steatohepatitis (NASH) is a chronic liver disease. There is a clear need to develop pharmacological treatment for patients with NASH as well as biomarkers that can diagnose the disease. We describe a trial of semaglutide treatment for NASH, identify key patient characteristics and compare the relationship of patient characteristics and non-invasive biomarkers/scores. NCT02970942 is a randomised, double-blind, placebo-controlled, multi-national Phase 2 trial of daily subcutaneous semaglutide (0.1 mg, 0.2 mg, 0.4 mg) in patients with biopsy-confirmed NASH, F1–F3 fibrosis, NAFLD Activity Score ≥ 4, and body mass index (BMI) > 25 kg/m2. Exploratory analyses were performed to evaluate correlations between baseline parameters and biomarkers in NASH. Mean (standard deviation SD) age of 320 randomised patients was 55 (11) years, mean BMI was 36 (6) kg/m2, and 199 (62%) had type 2 diabetes. Of the total patients, 28% had F1 fibrosis, 23% had F2 fibrosis and 49% had F3 fibrosis. The highest area under the receiver operating characteristic curve (0.69) for accuracy in classifying fibrosis stage, F2–3 versus F1, was observed for Fib-4 and Enhanced Liver Fibrosis (ELF). No substantial correlation between BMI or other clinical or biochemical parameters and fibrosis stage was observed. In this large Phase 2 trial of semaglutide treatment for NASH, the clinical profile of enrolled patients was typical for patients with NASH. Of the investigated biomarkers/scores, ELF and Fib-4 showed the most apparent correlation in classifying fibrosis stage, but had only moderate predictive value. •Liver fibrosis in patients with NASH is associated with disease complications.•Fib-4 and ELF were better able to classify fibrosis stage than other biomarkers.•BMI did not appear to be related to fibrosis stage.