Objective The qualities of intraoperative fluoroscopy images,which are the most important source data,have significant influence on the accuracy of computer assisted orthopedic surgery (CAOS) system.The sensitivity of the accuracy in four typical image quality parameters are studied based on the bi-planar positioning method.Methods Fiducial object with steel balls are used as the experiment object,and high quality reference images are acquired with high precision digital radiography device.Then the qualities of the images are quantitatively reduced by human intervention.The images with modified quality parameters are used as input data of bi-planar method to test the positioning error.Finally the Monte Carlo method is used to analyze the influence of random image error on the positioning accuracy.Results There is a significant correlation between the image resolution,distortion and positioning error,while image contrast and signal-to-noise ratio have no significant effect on positioning error.Monte Carlo analysis shows that random image error within ［-10,10］ pixels can lead to position error of 11.65 mm±9.06 mm.Conclusions The quality of the fluoroscopy images can significantly affect the spatial positioning accuracy,and the resolution and the image distortion have significant influence on the positioning error,while the contrast and signal-to-noise ratio are relatively small.A reasonable standard of intraoperative fluoroscopy image quality on the premise of ensuring the operational accuracy has a great significance for improving the positioning accuracy of CAOS system.