This paper focuses on the gendered choice of high school in the Italian context, where children are tracked at age 14 and are free to choose the type of school, with no binding teacher recommendation or ability restriction. It is therefore a context in which preferences, however influenced by different factors, are freely expressed, without any institutional constraints imposed on the decision-making process. Previous literature has mainly analysed gendered educational choices by focusing on the field at later stages in life. The transition from lower secondary to upper secondary school is particularly relevant for children who do not go on to university and can help to understand gender segregation in low and middle-level occupations. We analyse the role of school performance in mathematics and Italian (teacher grades and standardized test scores), the position in the class ranking, the comparative advantage in one subject and find that, while school performance hardly explains the gender gap for the children with low educated parents, it explains part of the gender gap observed for children from more advantaged backgrounds.