Abstract: The effects of the simplified camera model on 3D reconstruction from two images are investigated, the main conclusions are: When the camera motion between the two images is a pure translation, it is proved that the reconstructed object under the simplified camera model is an affine transformation of the original one. When the camera motion between the two images is a general rigid motion, some rather strict conditions must be satisfied in order for the reconstructed object under the simplified camera model to be faithful of its original shape. Under the simplified camera model, the estimated effective focal length through the Kruppa equation is in general not accurate enough for 3D reconstruction; a full camera calibration is needed for any real applications in 3D reconstruction, and the simplified model is in general not a viable replacement.