Objective The present work is an analysis of whether adoption of state clean indoor air laws (SCIALs) covering bars reduces the proportion of bartenders who smoke primarily by reducing smoking among people already employed as bartenders when restrictions are adopted or by changing the composition of the bartender workforce with respect to smoking behaviours.

Methods Logistic regressions were estimated for a variety of smoking outcomes, controlling for individual demographic characteristics, state economic characteristics, and state, year, and month fixed effects, using data on 1380 bartenders from the 1992–2007 Tobacco Use Supplement to the Current Population Survey combined with data on SCIALs from ImpacTeen.

Results State restrictions on smoking in bars are negatively associated with whether a bartender smokes, with a 1-point increase in restrictiveness (on a scale of 0–3) associated with a 5.3% reduction in the odds of smoking. Bar SCIALs are positively associated with the likelihood a bartender reports never having smoked cigarettes but not with the likelihood a bartender reports having been a former smoker.

Conclusion State clean indoor air laws covering bars appear to reduce smoking among bartenders primarily by changing the composition of the bartender workforce with respect to smoking rather than by reducing smoking among people already employed as bartenders when restrictions are adopted. Such laws may nonetheless be an important public health tool for reducing secondhand smoke.