Background/Aim Simple, inexpensive and clinically available noninvasive liver fibrosis tests are highly needed. We aimed to develop a novel noninvasive index for predicting significant fibrosis and cirrhosis in chronic hepatitis B (CHB) patients. Methods Using liver histology as gold standard, we developed a novel index to predict significant fibrosis and cirrhosis in CHB patients and then compared the diagnostic accuracy of the novel index, aspartate transaminase-to-platelet ratio index (APRI), and fibrosis index based on four factors (FIB-4) in a training set (606 patients) and a validation set (216 patients) from the same patient catchment area. Results Of 606 CHB patients in the training set, 33.2% had significant fibrosis and 11.4% had cirrhosis. In multivariable analysis, gamma-glutamyl transpeptidase (GGT) (OR=1.032, p<0.001) and albumin (OR=0.953, p=0.048) were independent predictors of significant fibrosis. Consequently, a GGT-to-albumin ratio (GAR) was developed. In the training set, the area under the receiver operating characteristic curve (AUROC) of GAR was significantly higher than that of APRI and FIB-4 to predict F2 (0.82, 0.70, and 0.68, respectively), F3 (0.86, 0.76, and 0.75, respectively), and F4 (0.88, 0.75, and 0.73, respectively), respectively. In the validation set, the AUROC of GAR was also better than APRI and FIB-4 for predicting F2 (0.81, 0.63 and 0.61, respectively), F3 (0.88, 0.78, and 0.76, respectively) and F4 (0.92, 0.85, and 0.78, respectively), respectively. Conclusions GAR is a more accurate noninvasive index than APRI and FIB-4 to stage significant fibrosis and cirrhosis in CHB patients and represents a novel noninvasive alternative to liver biopsy.