Abstract: When Global Navigation Satellite System (GNSS) deformation monitoring is applied to the railway infrastructure, it is difficult to maintain the railway line on site due to its continuous closure during operation. Therefore, it is necessary to analyze and evaluate the quality of GNSS observation data at GNSS monitoring stations to check their operational status. Based on the analysis of commonly used GNSS observation data quality evaluation indicators and combined with the characteristics of deformation monitoring, this paper selected 9 evaluation indicators, used ridge regression analysis, established a prediction model between multiple data quality indicators and the final calculation accuracy of monitoring, and conducted absolute evaluation of data quality. Based on the observation data of 8 GNSS monitoring stations on a slope along the Haoji (Haole Baoji - Ji'an) railway, the paper established a prediction model. After verification, the root mean square of the prediction error of the model was 1.1 mm, which was superior to the conventional error of GNSS solution. This proves the reliability of the model and the feasibility of the evaluation method.