Average population density has been a traditional urban sprawl indicator, but it is insufficient to encapsulate its detrimental impacts. We demonstrate that cities of identical average population density may be subject to very different levels of car dependency, CO2 emissions and public budget deficits. To capture this, we shed light into the properties of two “density-allocation” indicators: the percentage of urban land in which population density falls below a predetermined threshold value, and the fraction of the population living in such areas. Using unique data, we measure these indicators in more than 1100 urban areas of 29 OECD countries and at three time points: 1990, 2000 and 2014. We identify four different patterns for the coevolution of average population density and these indicators. The results show that densification of urban cores concurs with rapid suburbanization in many cities. We perform intertemporal cross-city and cross-country analyses and discuss the multiple policy implications emerging from them. •We examine the properties and socioeconomic importance of density-allocation sprawl indicators.•We measure the density-allocation indicators in more than 1100 cities, 29 OECD countries and in three time points: 1990, 2000, 2014.•We identify four patterns for the ways density-allocation indicators coevolve with average population density.•Reducing sprawl through changes in the density-allocation indicators involves different policies than those being effective in increasing average population density.•Core densification may concur with rapid suburbanization.