Cities are now in the spotlight for being responsible for a predominant share of anthropogenic environmental pressures. Urban metabolism is a conceptual tool that is increasingly used to monitor resource and energy use as well as pollution flows entering and exiting cities. Recently, effort has been put in the creation of consistent methodologies to measure the metabolism of cities in order to enhance comparability. Such exercises allow for cross-city comparisons at the macro-scale and have been illustrating potential factors explaining macro-scale differences between cities. However, such studies often rely on very few data points - just a handful of cities - and they look at cities as homogenous entities. When looking at the relationships between urban characteristics and metabolic flows at smaller spatial scales (at municipality or district level), the drivers appear to be different than in macroscopic analyses, pointing towards the importance of taking microscale urban heterogeneity into account. The aim of this paper is to improve our understanding of these effects. Data on energy use and various urban indicators was collected at a microscale level for ten cities (Brussels, Buenos Aires, Cape Town, Chicago, Glasgow, London, Los Angeles, Milan, New York City and San Francisco). It was then analyzed whether the effects which had been observed at the macro-scale (economies of scale, affluence effects etc.) are also valid if we look at cities in more detail. Our findings have shown that the additional detail gained by looking at micro-scale data questions some of the insights gained by analyses done with aggregated urban data. Results show that although the relationships and general trends found by previous studies looking at macroscale data also emerged in our study, the relationships are much less clear and that there is a high degree of heterogeneity across the cities, pointing towards the fact that to truly understand and manage energy use in urban systems, it is necessary to look at microscale level characteristics.