Traffic demand analysis and prediction are important for urban rail transit planning. However, the complexity of macro traffic model, data quality and planning changes are easy to lead to the uncertainty of ridership prediction. With Shijiazhuang taken as an example, this study employs mobile phone big data to analyze urban characteristics based on the four dimensions of job-housing distribution, job-housing balance, travel demand, and commuting circle to compensate for the uncertainty in traffic demand and ridership predictions. The data is also used to compensate for deficiencies in the demand prediction model in the three aspects of urban feature description, current traffic model verification, and development trend evaluation. Finally, the results are compared with those of Paris Region (Région ?le-de-France); drawing on the experience of rail transit planning and development in Paris Region, the study provides a decision-making reference for urban rail transit planning.