The Northern Finland Birth Cohort 1986 (N = 8,061) was used to investigated the joint associations of physical activity and sedentary behavior with academic achievement. Logistic regression analysis was used to evaluate how classes formed by latent class analysis (LCA) according to overall physical activity, sports club membership, viewing TV, using a computer, reading books and magazines, other sedentary activities, and sleep were associated with grade‐point average. When adjusted for gender, self‐rated health, and mother's education, physically active adolescents and generally active adolescents were about twice as likely to have high grade‐point average compared with sedentary TV viewers.