Affective disorders are pathological alterations of phylogenetically old emotional mechanisms the adaptive function of which is still a matter of debate. According to the social competition hypothesis, depressive syndromes involve activation of a mechanism serving subordination in competitive situations. This mechanism protects the individual from the damaging consequences of an escalating agonistic encounter. The paper critically discusses supporting evidence culled from biological, clinical and psychological findings. Animal models show phenomenological similarities between subordination and depression. In primates affiliative behavior and social bonding are closely linked to rank-related behaviors. Common biochemical and pharmacological features of subordination and depression include an activation of the stress axis, diminished serotonergic neurotransmission and modification by antidepressants. However, the discrimination between a specific subordination reaction and an unspecific stress response is often controversial. Many clinical and psychological aspects of depressive syndromes may be understood as an--exaggerated--submissive behavior, but have rarely been studied in that light. Typical endogenous depression and bipolar disorders challenge the social competition hypothesis. While the manic syndrome, due to its similarity to dominance behavior, supports the link between social competition and affective disorders, the cyclic alteration of mood cannot be sufficiently explained by the social competition hypothesis alone.