Objective To evaluate whether the adoption of a local clean indoor air (CIA) policy in St. Paul, Minnesota, was associated with changes in alcohol-related crimes outside on-premises alcohol-licensed businesses.

Design The enactment of a comprehensive CIA policy on 31 March 2006 was used as the intervention time point in an interrupted time-series analysis to assess changes in weekly crime frequency prior to the policy enactment compared with the period after the policy was established (n=261 weeks).

Setting St. Paul, Minnesota, USA.

Subjects On-premise alcohol-licensed business addresses were collected from St. Paul, Minnesota, for the period of January 2003 to December 2007, and geocoded. A 500-foot (152.4 m) buffer was drawn around each business.

Main outcome measures Alcohol-related crime (ie, arrest) data were obtained from the St. Paul Police Department; crimes had been geocoded by the police department. They were aggregated by week to include only those crimes that occurred within the drawn buffer. Relevant types of crimes included serious (eg, aggravated assaults, homicide, robbery, rape and theft) and less serious (eg, lesser assault, fighting, noise violations, public drunkenness/lewdness or other liquor law violations) crimes.

Results Within a buffer of 500 foot of alcohol-licensed businesses, 23 978 serious alcohol-related crimes and 49 560 less serious alcohol-related crimes occurred over 5 years. Using interrupted time-series analyses to compare the weekly alcohol-related crime frequency in proximity with the bars and restaurants, we found no significant change in either type of crime associated with the local comprehensive CIA policy (p=0.13) after adjustment for seasonal differences and overall crime frequencies.

Conclusions Evidence from this study suggests that alcohol-related crimes were not significantly affected by a local comprehensive CIA policy that banned smoking in public workplaces in St. Paul, Minnesota.