This paper proposes a new method to evaluate the reliability of published empirical formulas in terms of accuracy and applicability to different soil types. Different empirical models are proposed to properly approximate the compression index for a wide range of water contents and soil types. They were developed using a unique technique and a substantial number of published regression equations and compression data. Familiar empirical equations were examined for their reliability in predicting the compression index of clay for any water content. A comparison was made between available and newly-proposed empirical formulas using combined regression data sets compiled independently by several authors. The newly proposed empirical compression index equations are applicable to a wide range of clay soils, and in validating other published relationships. The degree of scatter and variations in the computed compression index values are minimized for any water content.