The study is based on two waves of longitudinal data at ages 13 and 16 aimed to examine factors that explain psychosomatic and depressive symptoms at the age of 16. The sample is nationally representative involving 9000 individuals from the Swedish longitudinal Evaluation Through Follow up database born in 1998. A growth modeling approach was applied to examine relations between factors over time. Girls were found to have considerably higher self-reported levels of mental health problems in grade 9 than boys. The gender difference was partially mediated by girls' higher level of stress and worry in grade 6 and larger increases in stress, worry and academic demands from grade 6 to 9. The larger increases in girls' feelings of peer exclusion were associated with the self-reported mental illness measures. Students with high cognitive ability experienced increased worry and stress in grade 9 with associated mental health problems.