There has been a decade-long debate in the studies of Malaysian politics on whether there is indeed an urban–rural difference when it comes to elections. Studies using aggregated election data suggest stark differences in parties’ performance in urban and rural electoral districts, while studies relying on survey data tend to downplay urban–rural differences in voting patterns. Notwithstanding the ecological fallacy problem inherent in studies using aggregated election data, the consistent differences between studies using individual and aggregated data are puzzling, and cast a shadow over our understanding of electoral politics in Malaysia. This article argues that in Peninsular Malaysia the urban–rural differences supported by aggregated election data may have been overestimated due to results being driven by a few large urban centers. Combining survey data from the fifth wave of Asian Barometer and aggregated election data from the fourteenth general election in Malaysia, this article demonstrates that both kinds of data in fact point to the same conclusion. Once we specifically control for inter-state and local heterogeneity in population density, the association between population density and party performance attenuates.