Background: A great interest has been dedicated to the development of non invasive predictive models in recent years to substitute liver biopsy for fibrosis assessment and follow-up. The aim of this study was to comparethe accuracy between S-index and aspartate aminotransferase to platelet ratio index (APRI) to FibroScan for predicting liver fibrosis in chronic hepatitis B and C patients.Method: A cross-sectional study was conducted in 40 patients with chronic hepatitis B and C between January 2010 - May 2011 at Division of Gastroentero-hepatology, Department of Internal Medicine, Adam Malik Hospital, Medan. Patients underwent laboratory examination and FibroScan, then used predictive values to assess the accuracy of S-index scores and APRI compared to FibroScan. The analysis was performed using SPSS 15.0.Results: S-index identified significant fibrosis in 87.5% patients with sensitivity (Se) 87.5% and specificity (Sp) 100%. About 67.5% of 40 patients could be identified correctly. S-index also could accurately predict the absence or presence of cirrhosis in 87.5% of the total 40 patients, with NPV 91.7% and PPV 81.25%, respectively. APRI for significant fibrosis has Se 85.7%, Sp 88%, PPV 88.8%, NPV 69.2%; while Se 53%, Sp 88%, PPV 72.7%, NPV 75.8% for liver cirrhosis. AUROC value for S-index was higher than APRI in predicting significant fibrosis and cirrhosis, i.e. 0.938 vs. 0.917 and 0.873 and 0.707, respectively.Conclusion: The S-index has a higher accuracy than APRI in predicting significant fibrosis and cirrhosis in patients with chronic hepatitis B virus and hepatitis C virus infection.Keywords: S-index, APRI score, FibroScan, chronic hepatitis B and C