We analyse the academic performance of Italian students who graduated in 2004, and their occupational status and earnings in 2007. We find that the educational and occupational performances of male and female students do differ: girls outperform boys in academic achievement, but male graduates outperform female graduates in labour market outcomes. One could wonder why female students put more effort into educational performance than male students, given that they will receive lower wages. We find a rationale for this choice in the higher marginal return that female students gain from their higher grades. We address our empirical analysis to four points: first, we show that, for the most part, the difference in educational performance is explained by the diversity in unobserved characteristics between male and female students. Second, we provide empirical evidence that the amount of effort supplied is the key determinant of the unobserved characteristics. Third, we argue that female students study hardly because they gain a higher marginal return from success in educational competition. Fourth, as this finding may be consistent with both human capital and sorting models of education, we test the hypothesis that female students use their higher grades to signal their ability to potential employers.