Abstract: Strike-slip fault is a common geologic hazard for buried pipelines. The fault-induced ground displacement often leads to excessive deformation and failure of pipelines, so it is of great significance to carry out reliability analysis of pipelines in fault areas for safety assessment of pipelines. Therefore, the finite element model of X80 steel pipeline crossing strike-slip fault was established based on the nonlinear finite element software ABAQUS, the design strain of the pipeline under the influence of many factors such as pipeline geometric dimensioning, surface displacement, inner pressure and soil type was calculated, and the finite element data of the pipeline design strain in the project was formed. Based on this database, the design strain prediction model of double hidden layered BP neural network was built, and the limit state equation based on strain criterion was established. Combined with MC (Monte Carlo) method, the reliability calculation of X80 pipeline in strike-slip fault area and the analysis of influencing factors of pipeline reliability were conducted. The calculation result based on BP-MC method is accurate and the running time cost is low, so it will be applicable to the reliability analysis of pipelines under faulting.