Abstract: The internal rock mass is widely distributed with structural planes,and its shear strength exhibits nonlinear characteristics under different normal stresses,influenced by joint roughness coefficient(JRC),rock strength(JCS),and residual friction angle(\varphi_\mathrmr).Based on the Mohr-Coulomb criterion,the traditional Morgenstern-Price method calculates normal stresses but fails to fully consider the variability of shear strength parameters of rock mass.This study introduces the Barton-Bandis criterion into the Morgenstern-Price method,proposing an improved Morgenstern-Price method(referred to as MP-BB method hereafter).The MP-BB method is then used to analyze the slope stability and calculate the failure probability of the slope in conjunction with the improved Rosenblueth method.The research shows that the MP-BB method takes into account of the variability of shear strength parameters of rock mass(JRC,JCS,\varphi_\mathrmr) associated with structural planes,fully reflecting the true statistical distribution characteristics of these parameters.The stability coefficient calculated based on the MP-BB method is more accurate than that of the M-C criterion.The correctness of the failure probability calculation results using the MP-BB method is verified using the Monte Carlo method.Therefore,the MP-BB method can be considered an effective slope stability analysis method.