Objective To build a practical and convenient diagnostic model of primary open-angle glaucoma （POAG） using the frequency-domain optical coherence tomography （FD-OCT） ethnic-specific database of normal humans for clinical application. Methods In a cross-sectional observational study， 133 eyes from 133 healthy subjects and 99 eyes from 99 early POAG patients were included in the study. MM 6/Radial slicer， optic nerve head （ONH）， retinal nerve fiber layer3.45 （RNFL 3.45） and ganglion cell complex （GCC） scans were measured in each subject with an RTVue-100 FD-OCT. Then， these parameters were used to establish the ″diagnostic models″. Receiver operating characteristic （ROC） curves and the area under the receiver operating characteristic curve （AROC） were used to reflect the capability of scores for each parameter of the FD-OCT in order to distinguish eyes in early-stage POAG from normal eyes. A combination of parameters was used with logistic regression to establish ″diagnostic models″ for early POAG. Results ①Significant differences were observed for all grading scores of ONH， RNFL3.45 and GCC scanning parameters between normal and early POAG groups （P＜0.05）. ②The best single parameter for ONH scanning was the RNFL average （R-A） （AROC， 0.912）. ③The AROCs of the ″diagnosis model″ were greater than the best single parameter. Their AROCs had significant differences. ④We validated these models. The AROC of the ″diagnosis model″ in ONH scanning was 0.987 and was 0.959 for RNFL3.45 scanning. Conclusion A multi-parameter ″diagnosis model″ using FD-OCT parameters and grading based on an ethnic-specific database of normal humans can improve the ability to diagnose early POAG. It may be convenient for clinical application.