Compared to a single camera, multi-camera systems view a larger field and collect more data for motion estimation and 3D reconstruction. The constant rigid body transformation between the cameras mounted (mounting parameters) is valuable knowledge for outlier detection and estimation of the system motion in real scale. In this research, the mounting parameters of the cameras are estimated in single step through bundle adjustment. Taking advantage of the mounting parameters, a methodology is proposed to estimate the system motion between successive data acquisition epochs. In this method, several estimates for the system rotation are obtained by constraining the relative rotation of the individual cameras. The outliers within the rotation estimates are filtered out and the inliers are averaged. Then, the system translation is estimated by solving a system of linear equations. The experimental results show that the proposed methodology can successfully recover the multi-camera system motion.