There has been much debate over the last three decades on the relationship between design of housing and crime, dominated by Newman's 'defensible space' and Coleman's 'antidisadvantagement scales'. Their approach has caused much confusion due to their failure to distinguish between social and spatial 'causes'. This thesis presents research conducted on the relationship between design of urban housing estates and spatial vulnerability to crime, addressing this important design issue, from both a theoretical and an empirical perspective. It examines the multi-disciplinary discourse on housing design guidance; design against crime and criminological insights on the spatial factors influencing the location of crime and selection of crime targets. It proposes a methodological framework applying the syntactic approach developed by Professor Bill Hillier and the former Unit of Architectural Studies at the Bartlett, University College London, through which the factors of spatial vulnerability can be identified and analysed in purely spatial and architectural terms. Based on a series of case studies on the locations of burgled dwellings in three London housing estates, the research identifies the two basic principles of vulnerability: accessibility and surveillability at the local level of articulation of space and the global level (of the spatial network as a whole), and the interrelationships between them, which ultimately determine the vulnerability of locations. Rather than condemning a list of block features, spatial layout is treated as a whole system, and the spatial factors of vulnerability are related to architectural choices at the various levels or stages of the design layout. Each estate design has its own individual set of combinations of spatial variables and constants in the layout and block/dwelling typology related to the parameters outlined above, which lead to a trade off between the likelihood of being seen or caught and the difficulty of getting in and out. Ultimately, visibility and accessibility are related to 'surveillability' in two forms: visual and social (active presence of people). However, it is not just visibility that is important, but also permeability (links via dwelling entrances), which allow direct active control over space (interception) The implications in terms of design for safer housing environments are, that simple recipes of good and bad design are highly questionable, since spatial configuration has to be examined as a whole. Thus understanding of the spatial principles and mechanisms of vulnerability as the intrinsic interrelationship between local and global factors, is the key to designing safer urban housing environments.