In promoting 'defensible space' as a crime prevention measure, Newman reasons that an environmental design which facilitates residents' territoriality and sense of community can be translated into residents' responsibility for ensuring a safe and well-maintained living space. The potential criminal will perceive such space as controlled by its residents and thus be deterred from attempting crimes in that space. In attempting to verify his defensible space model, Newman uses two general sources of information; records of crimes committed during 1969 in 100 New York City public housing projects and a comparison of two housing projects relatively identical in the number and density of tenants. Newman claims that the housing project with the lower crime rate has more defensible-space characteristics in its design and layout. Newman advances design features that foster territoriality, natural resident surveillance, and image and milieu. He tests these concepts with data on crime, fear of crime, and perceptions of community instability. The initial version of Newman's defensible space model ignored social phenomena as either predictor or mediating variables. The community-building rationale is important but raise methodological problems for evaluation research. The future of environmental design research is in illuminating the process by which people will come together to form a protective alliance. 70 notes and 10-item bibliography.