Underground pipelines inevitably go across rivers. For different crossed rivers, large differences may exist in their catchment areas, topography, rainfall, underlying conditions and so on, and then their flood threat will vary among corresponding infrastructure segments. Therefore, a simple but effective assessment method for flood threat is urgently needed. In this paper, the section of the Shaanxi-Beijing Gas Pipeline in Linxian County, Shanxi Province was chosen for case study. Firstly, flood factors were quantified for each crossed river, including the area, rainfall, elevation difference, planar shape, land use, vegetation index of each river catchment, and the longitudinal slope and cross-sectional elevation difference of each cross-over region. Secondly, three key factors were selected after the cross-correlations among the factors and the relationships between each factor and the number of historical flood damage events were examined. Thirdly, the multivariate linear regression and logistic regression methods were used to establish a flood threat assessment model. Finally, the flood threat zoning map of the Shaanxi-Beijing Gas Pipeline in Linxian County was obtained. The result would provide a basis for the formulation and implementation of protection measures in pipeline safety management, and the method in this paper could be a reference for similar flood threat assessments of linear infrastructures.