Information was collected from the Detroit Morning News, the Dallas Morning News, the San Francisco Chronicle, the Cleveland Plain Dealer, the Albany Times Union, and the Buffalo News. The sample included crime stories presented every 5th day during the first 6 months of 1990, although stories smaller than 3 column inches were excluded. The total number of specific incident stories was 1,557. Multiple regression techniques were used to determine the important variables that increase the salience of crime stories. Results revealed that the number of victims affected by a crime is the best predictor of increased story salience. The type of offense also had some influence. Other variables such as victim and defendant characteristics had very little influence on the space and attention provided to newspaper crime stories. Tables, notes, appended table, and 43 references (Author abstract modified)