In geometric positioning of optical line array remote sensing image, as it exists error on system error correction parameters corresponding coefficient matrix, a geo-positioning method of remote sensing image of optical line array based on regularized total least squares was proposed in this paper. Firstly, the collinear equation was constructed based on the definition of rational function model, Secondly, the geo-positioning and systematic error correction methods were derived by linearization. Then, the corresponding optimal objective function was constructed based on the definition of the EIV model with a regularization term introduced. The iterative solution method of system error parameters based on regularized total least squares solver was deduced by the Lagrange conditional extremum. Experiment results indicate that, the geo-positioning accuracy of this method is increased by 11.61% compared with the classical least squares adjustment algorithm, and it is increased by 6.06% in contrast to Tikhonov regularization geo-positioning accuracy. The proposed method in this paper can increase the geo-positioning accuracy of linear array CCD images effectively without any extra information.