Currently, office vacancy rates in the Netherlands are reaching record breaking levels. Seven million square metres office space is vacant; approximately 14% of a total stock of 47 million square metres. Given the declining demand for office space combined with the impediments related to reducing the office stock, this vacancy level is only expected to rise. The current oversupply on the office market results from both market imperfections and a failing planning system, only partly explained by the well known effects of the hog cycle. A more important explanation is the overproduction of offices that has taken place in the last decade, fuelled by low interest rates and high returns, sustained by a municipal will to develop land, and made possible by governmental belief in free market forces. The emergence of a replacement market, allegedly a market in which good buildings drive out bad buildings (Geraedts and Van der Voordt, 2003), has resulted in a steadily growing collection of obsolete office buildings. Supposedly, these buildings share features that are recognised as being related to obsolescence and depreciation (Baum, 1993; Bottom, 1998; Rem¯y, 2010). Characterising superfluous office buildings is however a difficult task as the odds of vacancy seems to differ along with their location or submarket. Vacancy levels are high on locations that are undesirable to tenants, either as a result of negative location externalities, or because the quality of the local office stock is lacking. In this paper, results from case studies of the Amsterdam and Haarlemmermeer office markets are discussed, showing that the features that can be used to describe obsolete office buildings is to a certain degree linked to local externalities.