A variety of approaches to delineatemetropolitan areas have been developed. Systematiccomparisons of these approaches in terms of the urbanlandscape that they generate are however few. Our paper aimsto fill this gap. We focus on Indonesia and make use of theavailability of data on commuting flows, remotely-sensednighttime lights, and spatially fine-grained population, toconstruct metropolitan areas using the different approachesthat have been developed in the literature. We find that themaps and characteristics of Indonesia’s urban landscape varysubstantially depending on the approach used. Moreover,combining information on the metro areas generated by thedifferent approaches with detailed micro-data fromIndonesia’s national labor force survey, we show that theestimated size of the agglomeration wage premium dependsnontrivially on the approach used to define metropolitan areas.