Abstract: [Objective] The stability analysis of rock slope is the basis to ensure the normal construction and safe operation of engineering facilities. Once the slope instability occurs, it will cause heavy casualties and economic losses. A large number of determining and random discontinuities widely developed within rock slopes lead to the discontinuity and heterogeneity features of rock mass, and have a significant impact on slope’s stability, deformation characteristics and failure mode. Hence, it is necessary to deeply research the development characteristics of discontinuities and the influence on the stability of rock slopes, which has great theoretical significance and practical value for slope protection and avoiding disaster losses. However, in current studies, the determining discontinuities and random discontinuity networks are rarely taken into account simultaneously to construct rock slope models and to research the effect on the probabilistic stability analysis and failure mechanism of rock slopes. [Methods] In this study, firstly, a three-dimensional numerical slope model is constructed using Rhino software, and the DFN (Discrete Fracture Network) model is applied to generate both determining discontinuities and random discontinuity networks on the basis of the field discontinuity data. Then, the coupling analysis model of determining and random discontinuities was established by integrating the slope model with determining discontinuities and random discontinuity networks within 3DEC software. Afterwards, the probabilistic stability analysis method of rock slopes was proposed to analyze the effect of random discontinuities on slope’s stability. Lastly, to illustrate the effectiveness of the proposed method, a simplified rock slope model is used to assess the impact of random discontinuity networks on slope stability. Additionally, the probabilistic stability analysis of the left bank shoulder slope of Jinping I Hydropower Station is conducted under the natural and excavated slope conditions. [Results]The results indicate that: (1) the proposed method not only effectively and accurately simulates the deterministic and random discontinuity network, but also realizes the coupling modeling of the discontinuity network and rock slope model. (2) The probability distribution diagram of slope stability coefficient can be obtained by the proposed method, and the calculated results are more comprehensive and in line with engineering practice. (3) The random discontinuity network has a more significant influence on slope stability under the excavated slope condition and changes the failure path and instability mechanism of the rock slope. [Conclusion]The research results not only provide references for the excavation and support scheme of rock slopes in engineering practice, but also provide a theoretical basis for the prevention and control of geological disasters.