Although negative self-evaluation is a common symptom of depression in adolescents, there is little understanding of how the self is associated with depression. Beck (1967) proposed that a negative view of the self was a 'hallmark' of depression. In contrast Linville (1985; 1987) proposed that holding multiple aspects of the self was associated with lower levels of depression. The aim of this paper is to evaluate these two models of self and depression in adolescents.Young people aged 13-18 years (n = 822) reported symptoms of depression (the Mood and Feelings Questionnaire) and completed a measure of self-concept, the Twenty Statements Test (TST). We coded responses to the TST to reflect the valance (positive to negative) and the complexity of their self-concept (number of self-aspects).Valence, but not complexity, of self-concept was significantly associated with severity of depression symptoms. The valance of young people's self-concept accounted for 25% of the variance in depression symptoms. Adolescent's with more positive self-concept tended to have fewer symptoms of depression.The cross-sectional design of this study precludes any conclusions about the causal relationship between depression and negative self-evaluation; experimental and longitudinal research is needed to assess the causal direction of the relationship.The results of this study supported the cognitive model of depression. Negative self-evaluation may constitute a risk factor for depression in adolescents and could offer a potential target for prevention and early intervention in adolescents.