Background: Mac-2 binding protein glycosylation isomer (M2BPGi) has been established as a non-invasive biomarker for liver fibrosis. We evaluated the diagnostic efficacy of M2BPGi compared with those of other liver fibrosis markers in liver fibrosis in non-alcoholic fatty liver disease (NAFLD). Methods: We analyzed serum M2BPGi levels in 113 NAFLD patients. A pathologist graded liver fibrosis histopathologically. The diagnostic efficacies of serum M2BPGi and other liver fibrosis markers (aspartate aminotransferase to platelet ratio index, fibrosis index based on four factors, and NAFLD fibrosis score [NFS]) were evaluated using correlation, area under the ROC curve (AUC), logistic regression, and C-statistics. Results: Serum M2BPGi level and other liver fibrosis markers showed a moderate correlation with fibrosis grade. The AUC values of M2BPGi were 0.761, 0.819, 0.866, and 0.900 for diagnosing fibrosis (F) > 0, F > 1, F > 2, and F > 3, respectively. Logistic regression analysis showed M2BPGi as the only independent factor associated with F > 2 and F > 3. Although C-statistics showed that NFS was the best diagnostic factor for F > 2 and F > 3, M2BPGi with NFS had an increased C-statistics value, indicating that it is a better diagnostic model. Conclusions The serum M2BPGi level increased with liver fibrosis severity and could be a good biomarker for diagnosing advanced fibrosis and cirrhosis in NAFLD patients. A well-controlled, prospective study with a larger sample size is needed to validate the diagnostic power of M2BPGi and other fibrosis markers in NAFLD.