Urban density measures are broadly used in urban planning as tools for calculating development yields, defining indicative targets, or development controls. A century ago Raymond Unwin argued in the influential pamphlet "Nothing Gained by Overcrowding" that limiting the number of dwellings per land area would be the simplest and best way to avoid overcrowding of both buildings and people, a measure which became widespread. The same measure was used by Jane Jacobs when describing the minimum density required for a vital neighbourhood. This density measure is however highly problematic as it does not account for the size of dwellings, mix of uses or size of households, factors that vary significantly in time and space. This implies that the number of dwellings per hectare alone is not useful for historic comparisons, urban versus suburban comparisons, or comparisons between different social contexts.
Likewise, pedestrian flow surveys conducted in three urban areas from Melbourne show that dwelling density is not a reliable indicator of streetlife intensity. Furthermore, it is shown that its vagueness makes it particularly vulnerable to be linked to any type of qualitative conception of density, thus contributing to the ongoing confusion and misuse in the density debate. It is concluded that dwelling density on its own is not an adequate measure to describe urban environments, and instead multi-variable density models need to be used in both research and urban planning. Such a model is illustrated for a range of urban morphologies.