Abstract: The majority of contemporary numerical model calibration methods focus on the usage of uniaxial compression tests, however the results of numerical models calibrated in this manner are generally inconsistent with the results of actual physical experiments with regards to the tensile strength and strength envelope. Therefore, based on the uniaxial compression test, the uniaxial tensile test and the triaxial compression test, the orthogonal experimental design method was used to study the calibration method of the mesoscopic parameters of the parallel bonding model. The multi-factor analysis of variance was used to determine the degree of influence of the model meso-parameters on the macro-mechanical indexes, and the relationship between the macro-mechanical parameters and the meso-parameters was established. On this basis, a determination method of meso-parameter matching with rock elastic modulus, Poisson′s ratio, uniaxial compressive strength, crack initiation strength, tensile-compression strength ratio and friction angle was proposed. The research results show that the macroscopic mechanical parameters and the uniaxial compressive stress-strain curve of the numerical model established by the calibration method in this paper are similar to the laboratory test results, which can well reflect the failure mode of the rock under uniaxial and triaxial conditions. The PFC2D fits the experimental limit principal stress curve poorly, which may be due to the fact that the parallel bonding model idealizes the irregularly shaped mineral grains into round grains, and the grain shape may affect the inter-grain occlusion.