ABSTRACT It is pointed out firstly that Fox’s method of computing the settlement of an embedded foundation is unsuitable because the solution given by Fox based on the superposition of Mindlin’s first solution does by no means conform to the stress state which can be visualized within the soil mass around a foundation. It can be concluded accordingly that the settlement thus computed is in general greater than the observed settlement, and especially when the depth of a foundation is large the degree of error will be excessive. Secondly, a new settlement formula is proposed by considering such factors as the net foundation pressure, the modulus of deformation and the bearing capacity below the foundation level which produce an essential effect on the settlement behavior of an embedded foundation.