Scholarship on the relationship between social structure and electoral behaviour has traditionally operationalised voters’ economic or class situations in terms of their labour market positions. Such an approach overlooks the growing importance of housing as a source of both socio-structural and electoral division. Using national election studies and household panel data from eleven countries, I conduct the first systematic, cross-national, and longitudinal analysis of the electoral behaviour of housing market groups, finding that housing represents a strong and growing electoral divide in advanced democracies. Electoral support for centre-right and ‘new left’ (that is, radical left and Green) parties is most strongly associated with housing market position, with homeowners significantly and increasingly more likely to support the former and renters significantly and increasingly more likely to support the latter. When other factors, such as house value and rental contract type are considered, housing-based electoral divides appear even greater in scope, to the extent that housing market position now appears a stronger predictor of voting for some party families – the centre-right and new left in particular – than labour market position. These findings challenge the claim that class or economic circumstances have become increasingly unimportant in structuring voting behaviour, and suggest an updated conceptualisation and operationalisation of social stratification – which moves beyond occupation and labour market position – is required to understand contemporary patterns of electoral alignment.