Issue preferences - and crucially priorities - are central to voting behaviour. To understand issue priorities, researchers often rely on the 'most important issue' (MII) question in political surveys, analysing voters’ issue priorities on aggregate and by different demographic groupings. Prior studies have used variations in responses to the MII question to explore gender gaps in issue salience and position and their relationship with voting behaviour (Campbell, 2012, Yildirim, 2022). However, the current measurement of MII questions presents limitations in capturing nuanced insights into the issue priorities of women and men. Firstly, MII categories are often broadly conceived, encompassing a variety of diverse and specific concerns into singular, overarching issue domains. Consequently, if men and women prioritise different aspects of the same broader issue, these differences are obscured. Secondly, standard coding of MII categories typically lacks measurement of issue position, potentially conflating vastly different voter attitudes towards an issue within a single category. This paper addresses these methodological limitations by employing quantitative computational text analysis techniques to a large corpus of open-text responses to the MII question. Our data sources include the British Election Study Internet Panel (2014-2024), British Election Study Face to Face Post-Election Surveys (2015, 2017, 2019, and 2024), and European Election Study (2019). By analysing differences in salience and position both within and across issues, we provide a more detailed understanding of men and women's issue preferences and priorities and offer novel insights into gendered voting behaviour.