The relative effects of various governmental interventions upon cigarette consumption is important to policy-makers. Historically, the demand for cigarettes has been quite unstable. Previous studies employ fixed parameter models and use dummy variables associated with interventions to stabilize the demand function. In contrast, we use a varying parameter model applied to data from the United States for 1953-84 to investigate the stability of demand and show that the demand function is stabilized when dummy variables are employed. Our results suggest that industry advertising increases aggregate consumption while government interventions decrease it. However, the marginal effect of government warnings seems to be small, at least in the US: while the effect of the 1964 health warning is statistically significant, the effect of the 1979 health warning is not. © 1991, Taylor & Francis Group, LLC. All rights reserved.