The mathematical model employed here to provide a more rigorous method to identify areas by CRD draws upon elements of models devised for the prediction of the incidence of both infectious and noninfectious diseases. First, the total area (a high CRD metropolitan area) is subdivided and the locations of the incidents are entered in a form recognizable by the computer. To this end the grid formed by the streets can be used, and the location entered by giving the horizontal and vertical positions in units of blocks. The areas of greatest CRD are located by moving a circle with a 1-block radius over the area of interest. Its center is successively placed at the intersection points of a search grid whose squares are two-tenths of a block on an edge. Then, the CRD within the circle at each search grid point is computed and compared to the average value of this quality. The best results are obtained with the search grid smaller than the street grid by a factor of at least three. This allows for a circle large enough to contain several incidents, but not including too large a fraction of the total number. This verifiable technique is more accurate than the traditional methods which rely on career experiences and visual inspection to interpret pin maps and even most computer-printed crime plottings. It also develops the foundation for distinguishing between crime density and crime seriousness. The question of crime seriousness factors still needs to be addressed. Figures illustrating the technique, footnotes and a bibliography are included. (Author abstract modified)