This paper considers the case of Italy to analyze the short- and medium-term effect of a longer school day in primary school on both students' learning and mothers' labor supply. we rely on unique application-to-primary-school data: first, we control for parental preferences, proxied by individual applications; second, we exploit variation in the probability of attending the full-time (FT) scheme that only stems from nonlinearities in the mix of FT and part-time (PT) applications received by the school and from class size limits set by the law. We show that attending the FT scheme increases Math test scores in grades 2 and 5 and Italian scores in grade 2 by around 4.5% of a standard deviation, but the effects fade away by grade 8. Conversely, there is a positive impact on maternal labor force participation and employment, which is long-lasting (approximately 2 p.p.). No effect is found on fathers' employment. Finally, we find some evidence of negative selection on gains, as the groups of students and mothers for whom the effect seems to be larger are not those more likely to apply to the FT scheme or to attend it conditional on applying.