Analyses indicated that there was a significant increase in the number of DUI stops following the implementation of the new closing time. Public officials and police attributed the increase to a variety of factors, including increased binge-drinking, more out-of-state drinkers and drivers, and changes in police practice. Subsequent bivariate analyses suggested that there were no significant differences in the characteristics of individuals stopped by police for DUI between the pre- and post-intervention periods. Similarly, there was no significant difference in blood alcohol levels, suggesting no change in levels of drunkenness as detected by police over this time period. Additionally, police were no more likely to use the most serious DUI charges. These result patterns suggested that the city did see a significant increase in the number of stops occurring during the hour immediately following the extended closing time, between 2:00 and 3:00 a.m., which pointed to differences in police practice, namely more proactive responses by police, rather than increased drunk driving. In anticipation of increased problems due to extended bar hours, local police administrators instituted patrol changes impacting evening and midnight shifts. Patrols during this time were more proactive, initiating more DUI stops, using various sobriety tests more often. This study used data on the number of police stops for DUI in one city in Minnesota over a period of 4 years from January 1, 2002, to December 31, 2005. Individual and incident-level data were collected for a subset of DUI cases occurring in Minnesota between January and December 2003, after closing times were extended. These two data sources and analytic strategies are described in detail. Figure, tables, notes, and references