Increasing complexity in existing and future engineering systems has led to the wide use in engineering practice of indirect measurements, i. e. , determining parameters characterizing an object by measuring primary parameters x//1, x//2, . . . , x//n functionally related to the unknown parameter Y and calculating the latter from the given function Y equals phi (x//1, x//2. . . . ,x//n). As in any measurement there are errors of random type, it is necessary to evaluate the error in determining Y, where one can use the mathematical expectation M// DELTA //y and the variance or dispersion D// DELTA //y equals D//y (the standard deviation sigma // DELTA //y equals sigma //y) of Y from the true value. This article discusses some techniques to handle this type of measurement.