BackgroundNegative attentional biases and self-schemas have been implicated in the development of depression. Research has indicated that a larger late positive potential (LPP) to negative self-referential words is associated with depression—as well as a maternal history of depression, an indicator of risk. However, it is unclear whether the LPP to self-referential words predicts the actual development of depression. The present study examined whether electrocortical reactivity during self-referential processing predicts the development of depression across adolescence. MethodsThe sample consisted of 165 8 to 14-year-old girls with no lifetime history of a depressive disorder who completed the self-referential encoding task (SRET) while electroencephalography was recorded at a baseline assessment. Participants and their parent completed the Kiddie Schedule for Affective Disorders and Schizophrenia for School Aged Children at the baseline, 2-year, 4-year, and 6-year follow-up assessments. ResultsResults indicated that a larger LPP to negative self-referential words at baseline predicted an increased likelihood of developing chronic-intermittent depression (i.e., persistent and/or recurrent), but not non-chronic, single episode depression, across adolescence. In contrast, neither SRET recall biases nor the LPP to positive self-referential words predicted the development of either type of depression. ConclusionsThe present study suggests that electrocortical reactivity associated with a negative self-schema in late childhood predicts the development of a more pernicious subtype of depression across adolescence. Moreover, the present study highlights the importance of considering clinical course in the examination of biomarkers of risk for depression.