This article seeks to examine the area characteristics that act as determinants of area property crime levels, namely, incidence and prevalence. The crime figures are extracted from the 1984 British Crime Survey. Area characteristics are taken from the 1981 UK census. Initial exploratory analysis considers the non-Gaussian nature of the crime data, the statistical implications of this, and the transformations used to overcome these problems. In addition, possible regional and inner-city/non-inner-city variations are considered. The later stages move from simple individual correlations to multiple regression models. Three regression models are considered and the reasons for refining these are explored, with the results indicating that both area characteristics and regional influences play a role as determinants of the area crime level. In particular, population density and the area population age profile have significant roles to play. The conclusions support the recent revival of the application of ecological concepts in the analysis of crime levels.