An examination of mayoral recruitment in 19 major American cities from 1820 to 1995 finds that mayors are now more likely to have prior political experience, serve longer in office, and have previously served on city councils. They are also less likely to come from business backgrounds, and those mayors who are still in business are less likely to have big business and manufacturing backgrounds. A multivariate analysis of mayors serving between 1870 and 1995 indicates that mayors from business backgrounds are less likely to be selected when their cities have partisan elections, when their cities have a large African American population, and when mayors are machine candidates. This analysis also finds that the number of mayors with business backgrounds declines as the number of college graduates increases, even though (after 1896) political machines disappeared and more cities adopted reform (council manager and commission) governments.During the colonial era American mayors were city council members selected by their councils to serve for short, usually one-year, terms. In the late 1820s mayors began to be directly elected and after the mid 19th century most mayoral terms increased from one to two and eventually four years. Power was consolidated in the 20th century, when city councils that were previously large and bicameral were reduced to single smaller chambers. These institutional changes are, however, only half the story. The other half of the story is how the kind of individuals who became mayor changed over time.As Robert Dahl (1961) taught us, the first question for any political system is who governs. Among the concerns of those interested in leadership selection is the extent to which business interests dominate government. When working with more limited historical data, the recruitment of mayors with business backgrounds is the best indicator of that relationship. Mayoral occupation is important because arguably it is associated with policy preferences and performance in office. Studies of public opinion discovered differences in opinion between the public sector, the manufacturing sector, and high-tech employees (Prysby, 1979). Studies of business faculty and students indicate that they have the most conservative opinions of any academic disciplines (Ladd & Lipset, 1975). Sidanius and Pratto's study (1993) of American and Swedish college students demonstrated a link in both societies between support for free-market capitalism and racism.