Despite the increases in women's educational attainment in recent decades, female labor market participation and labor market returns are still lower than those of their male counterparts. Among the main factors explaining this persistence of economic inequality is the persistently gendered nature of occupational expectations, which results in gender segregation of labor. In this paper, we describe how gender-specific adolescents' occupational expectations change over time (2006-2018) and how women's empowerment and cultural norms might influence gender-specific occupational expectations. Against the backdrop of the research on the gender-equality paradox and from a comparative perspective, we focus on national and institutional characteristics to investigate how individual and national factors explain gendered occupational expectations. We answer our research questions by applying a two-step multilevel model with fixed effects. For this, we used PISA data and merged them with state-level information from 26 European countries. We add to existing research by making three contributions. First, we describe the changes in occupational expectations over time within European countries by looking at the gender composition of the desired occupation and distinguishing three categories (gender-typical, gender-balanced, and gender-atypical). Second, we investigate the relationship between national characteristics and the evolution of gendered occupational expectations separately by gender to reveal gender-specific mechanisms. Third, by using data from two-time points, we explore which national-level changes lead to changes in students' occupational expectations. Our first descriptive results show that the patterns of how students' occupational expectations change over time differ remarkably between countries. In 2018 in some countries, students' occupational expectations became more segregated while in others the number of students with gender-balanced or gender-atypical expectations increased. Our fixed effects models show that women's empowerment and self-expression value explained variance over time. For example, women's empowerment measured via an increase in women's employment and participation in parliament led to less gender-typical occupational expectations among girls and boys. Similarly, a rise in self-expression values led to less gender-typical occupational expectations, again for both boys and girls. Remarkably, our results do not verify the gender-equality paradox for occupational expectations, as is the case in previous cross-sectional analyses.