How does the size of district magnitudes influence minority representation? Whilst this has been a classic topic in the institutionalist literature on representation, most causally robust evidence to date concerns close-list PR systems and the interactive effect of gender quotas (Lucardi and Micozzi 2010 AJPS), with the assumption that parties are the enforcer of these quotas. The literature has yet to causally examine the effect of district magnitudes on minority representation under less party-centred systems, such as open-list PR and SNTV systems, which are still prevalent throughout many democracies both at the national and subnational levels. This paper seeks to causally identify the effect of district magnitude size by employing a difference-in-differences design on an original dataset from Japanese local elections, which use SNTV with varying district magnitudes and where women are notoriously underrepresented. Since different levels (prefectural and municipal) of local elections are held simultaneously within identical district boundaries but often with different magnitudes, and exhibit only parochial differences in terms of candidate selection processes, the level of party-centredness and policy issues, Japanese local elections present a powerful case to find quasi-experimental evidence of any district magnitude size effect. As political experience at the local level is often a crucial step in political careers in many democracies, the findings will likely have implications on minority representation to numerous polities employing less party-centred electoral systems at the national or subnational level.