The use of satellite positioning techniques commonly requires a transformation from a Conventional Terrestrial coordinate system to a Geodetic coordinate system, or vice versa. For such a transformation, the main problem is the determination of transformation parameters between these coordinate systems. The transformation parameters are estimated by a least-squares process using "common" points, i.e., those points whose coordinates are known in both systems. Therefore, the precision of so estimated transformation parameters is closely related to certain characteristics of the common points. In this contribution, we have formulated some theoretical relations between the transformation parameters and the number and the distribution of common points, and corroborated the theoretical results numerically, using a simulated geodetic network.