Background Metabolic associated fatty liver disease (MAFLD) was recently proposed to replace non-alcoholic fatty liver disease (NAFLD), which is defined as ectopic fat deposition in the liver. Hepatic steatosis is an independent predictor for insulin resistance and cardiovascular risk and hence mortality. The aim of present study was to investigate the urine molecular pattern and potential urine biomarker to distinguish healthy control, mild and severe hepatic steatosis in MAFLD patients using proteomic data. Method Hepatic steatosis was measured by Magnetic resonance imaging (MRI) that measure the proton density fat fraction (MRI-PDFF). Proteomic measurements of urine samples were done by mass spectrometry. Linear regression analyses were used to detect significant associations between the biomarkers and clinical parameters. Western blot and Elisa were performed for validation. Results Multiple pathways have been compromised in MAFLD, including the carbohydrate derivative catabolic process, glycosaminoglycan process, aminoglycan metabolic process, inflammatory response, insulin-like growth factor receptor and GTPase complex. Alpha-1-acid glycoprotein 1 (ORM1) and ceruloplasmin were finally identified to be potential urine biomarkers to detect mild/severe hepatic steatosis.