We investigated the longitudinal relationship between physical fitness (flexibility, functional strength, and running speed-agility components) and subsequent change in academic achievement across one school year. We also examined whether this longitudinal relationship differed as a function of pupils' age, controlling for sex, body mass index, and socioeconomic status. Academic achievement in terms of marks in Portuguese and mathematics was recorded from 142 pupils (M = 14.59 years; SD = 1.99, range 11–18), between autumn 2017 and summer 2018. The physical fitness components, including flexibility, functional strength, and running speed-agility, were assessed at the baseline (i.e., at the beginning of the school year). Latent change score modelling revealed that higher physical fitness level at baseline significantly predicted a subsequent improvement in academic achievement across the school year. This longitudinal relationship was significantly stronger in younger compared to older pupils. Physical fitness and its interaction with age predicted 45.7% of the variance in the change in academic achievement. In conclusion, a better physical fitness profile including flexibility, functional strength, and running speed-agility explains a subsequent improvement in academic achievement. This longitudinal relationship seems to be age-dependent.