Women have caught up significantly with men in recent decades, but occupational choices remain stubbornly segregated by gender. One particularly popular hypothesis postulates that societal gender norms are the key factor explaining this phenomenon. I attempt to test the validity of this claim as directly as possible in this paper, using a unique combination of different data sources. Data covering the full stock of apprenticeship contracts active in the school year 2021/22 allows me to construct geographically disaggregated measures of occupational segregation, and I use local results from five national-level votes on gender-related policies to construct a measure of local gender norms. The data reveal that there is a statistically significant and practically relevant association between local gender norms and the regional extent of occupational segregation, with more progressive norms associated with less segregation. However, further analyses, carried out by leveraging two distinct identification strategies, point to a small and ultimately insignificant effect of gender norms on occupational segregation once potential confounders are taken into account.