The article presents the theoretical basis of the well-known method to detect and localise gross errors of measurements through correction for compensation. The results of the works in which it was applied have been summarised and complemented. The foundation of the technique was laid in the 1960s, yet, unfortunately, it has not been adequately publicised in the national geodetic literature, although today’s ways of the registration of measurements and calculations make the issue of the automated detection of possible gross errors very important. Highlighted are virtues and deficiencies of the method. The issue is significant for the land surveys when building roads, tunnels, and other construction objects.