Social defeat by Tryon Maze Dull rats, resulting in loss of rank of a previously dominant rat, has recently been advanced as a model of loss of self-esteem and depression in humans. However, before loss of rank in animals can be assessed as a model of human depression, certain requirements must be fulfilled: (1) a situation must be developed in which rank can be determined, and this assessment must, under normal circumstances, remain stable if it is to be the baseline against which experimental manipulations are measured; (2) it must be established that defeat of a dominant animal does indeed cause a change in rank for that animal; and (3) once these practical requirements have been fulfilled, to be a "model of depression," antidepressant treatment must be seen to reverse the loss of rank caused by defeat. Each of these points is considered in turn. (1) So far, a situation has not been discovered in which social hierarchies are produced that are both unequivocal and stable. (2) Defeat seems to have very variable effects, the variability being most likely due to natural variation in the quality of the defeat "experience" for the recipient. (3) Antidepressant treatment has not yet been shown to reverse the effects of defeat for more than a single test session. In addition, it has not been demonstrated that loss of rank actually produces a lowering of self-esteem in the animal as opposed to helplessness. If the change in self-perception does not change the value the animal attributes to itself, but rather the level of ability the animal attributes to itself, then the model breaks down. Consequently, at this time, the predictive validity of the model has not yet been established. Furthermore, its construct validity may be low. In summary, much more work is required before defeat-induced loss of status in the rat can be properly assessed as a model of loss of self-esteem and depression in humans, although it may be bedevilled by theoretical weaknesses even at this early stage.