Although criminological research has incorporated spatial variables as far back as the 19th century, professional geographers were not involved until the 1970's, when the behavioral dimension of criminality began to be extensively studied. Area studies of crime patterns in the 1970's mostly depended on cartographic procedures (such as centrographic and computer graphics techniques) and used official crime statistics of questionable accuracy. American studies often correlated urban zones with criminality, a technique not particularly applicable to British cities, where public-sector housing is a much stronger presence. Variants of ecological analysis correlated crime with environment; Oscar Newman's controversial 1972 study on defensible space was a benchmark in this area. Newman defined the concept of 'opportunities for crime,' and put blame on the anonymity of segments of urban space. Although many critics attacked the Newman study on methodological grounds, it launched numerous projects based on physical design as a crime deterrent. Data for two cities, Swansea, Wales, and Oklahoma City, Okla., are used to test several current hypotheses for the physical aspects of design, land use, location, and social dynamics of an urban area in relation to offense rates. Data on Cardiff, Wales, are used to test very different hypotheses concerning offender rates, which correlate offenders' homes, residential areas, and locality-based activities. Results suggest implications for research as well as the need for further refinement. An index and more than 180 references are supplied.