The study is unique in its reliance on survey data obtained from police regarding the severity of the local gang problem in jurisdictions in the south central United States. Questionnaires were mailed to 131 municipal police departments in Alabama, Arkansas, Louisiana, Mississippi, Missouri, Oklahoma, New Mexico, Tennessee, and Texas. This is a saturation sample of cities with a population of over 15,000 in these States. Over 60 percent of those surveyed (n=79) returned the two-page instrument, which had been addressed to the head of the gang control unit of each city's police department. The questionnaire addressed the influence of six independent variables on the perceived severity of the local gang problem: the number of gangs in the city, the size of the largest gang in the city, the predominant age range of the largest gang's members, the presence of out-of-town affiliations, the largest gang's longevity in the jurisdiction, and its organizational sophistication. Two identical regression analyses were used to explore the linkages between gang structure and police ratings of the severity of the gang problem at the local level. Data show that, at least for the region examined, police perceptions of the severity of the gang problem differ on the basis of jurisdiction size. So also do the predictors of that perception. The relatively low ratings given to perceived severity by a population whose interests would be served by overestimating the seriousness of the local gang problem contradicts the notion that street gangs are rapidly becoming a widespread urban problem. The data also provide a picture of differences in gang structures between large and small jurisdictions. The largest gangs in big cities are slightly better organized than those in smaller jurisdictions, and gangs are perceived by police to pose a more serious threat to public order in large cities as well. 3 tables and 37 references