This paper, which uses data on National Park visitors between 1979 and 1998 and every county in the United States, is the most exhaustive examination to date of how visitors affect crime. After controlling for many other factors that influence crime, the county-level regressions consistently indicate that national park visitors have no effect on either property or violent crime. These results are true for a variety of different measures of park visitors, for different empirical specifications, and for different regression formats. The authors therefore conclude that some visitor types have no impact on crime. This conclusion sheds light on the empirical issue of whether only some types of recreational visitors increase crime or whether visitors, regardless of their type, necessarily increase crime. (Published Abstract)