Data were obtained from the Adolescent Mental Health Study, a large-scale longitudinal study that involved five assessment points over a 1-year span. Two cohorts of sixth-graders and their parents were randomly selected from two suburban school districts in the Midwest United States, with a 1-year lag. Data assessments in the early adolescent years were conducted in the sixth, seventh, and eighth grades. Two follow-ups, each 4 years apart, were conducted, one during the 12th grade and one 4 years later. Adolescents were deemed at risk for the development of depressive symptoms if they had two or more of the following factors: insecure parental attachment, anxious/inhibited temperament, low instrumental coping skills, and early pubertal maturation. Depressive affect was assessed with a scale of the Self-Image Questionnaire for Young Adolescents (SIQYA). The emotional tone Scale of the SIQYA includes eight items that describe symptoms of depression. Growth curve modeling was applied to reveal significant gender differences in level and slope for the development of emotional tone over the adolescent years. At-risk girls revealed poorer levels of emotional tone that remained approximately at the same low level up to young adulthood, indicating a long-lasting impact of the specified risk factors. In contrast, at-risk boys showed poorer emotional tone in early adolescence, but at the end of adolescence boys at risk for depression had the same level of emotional tone as nonrisk boys. At-risk girls showed significantly higher levels of depressive symptoms at grade 12 and 4 years later. 2 tables, 3 figures, and 33 references