Musicians are trained in melodic transposition, the skill of extracting the pitch interval structure and moving it into different keys. This ability to recognize whether a melody is the same or altered when it is played back in a different key is correlated with both greater neural activation and cortical thickness in bilateral intraparietal sulcus (IPS). Musical training only explains part of this finding, suggesting that the ability to transpose a melody may have innate predispositions. The current study was designed to address this question: Are the anatomical correlates of musical transposition already present in non-musician children at age 14.5? If so, is there any evidence that those traits were already in place at earlier ages? To answer this question, we recruited 47 adolescents (age 14.5 years) from a longitudinal study and tested them on a melodic transposition task. These adolescents had already undergone anatomical MRI imaging at the ages of 10, 11.5, 13, and 14.5 years. They were tested on the transposition task at age 14.5. During this visit, we found a relationship between cortical thickness in left IPS and performance on the transposed melody task in the girls and not the boys; no such relationship was observed at any of the earlier ages. Girls reach more advanced staged of pubertal maturation earlier than boys, it is possible that the relationship between cortical thickness in IPS and skill at melodic transposition only emerges once the brain has reached a certain degree of maturity. This claim is supported by a lack of similar sex differences in the adults: the degree of correlation between cortical thickness and performance on the same transposed melody task did not differ between men and women in a previous study. Thus, our results suggest that the relationship between cortical thickness and the ability to transpose a melody is not fixed, and that the effects observed in adults are neither due exclusively to training nor to predisposition.