In this paper I propose a model of gentrification based on the notion that gentrification closes a gap between the flow of housing services fixed in a particular vintage of the housing stock and those available from the most modern properties. This gap is not a rent gap, therefore, but an investment gap. Modelled in this manner, gentrification appears as a problem of maximization under constraint and a subsubset of general home improvements. It is a transient and historically unique (noncyclical) phenomenon. Similarly, these constraints and the opportunities currently available to overcome them exist only in a particular historical context, the peculiarities of which must also be taken into account if gentrification is fully to be explained. In particular, these include the development of domestic technologies and the 19th-century conditions of supply of the housing available currently for gentrifying. I concentrate on the supply-side issues in gentrification, but deny that the demand-side issues can be handled via explanations based on postindustrialism, postmodernity, or the rise of a new middle class.