There has been a considerable debate about the need for stronger policies to restrict the advertizing of tobacco in the U.K. Medical bodies have argued for a complete ban but this has been opposed by tobacco companies and the Advertizing Association on the grounds that advertizing does not raise the level of tobacco consumption but instead only affects brand shares. Empirical studies have been cited by the two opposing sides in support of their positions. In this paper, some of these empirical studies are critically reviewed in order to throw some light on their limitations and the reasons for differences in results. The relevance of these studies in predicting the full consequences of an advertizing ban and other more general issues relating to the Government's policy on advertizing tobacco are also considered. Finally, some alternative advertizing policies available to the Government, and their relationship to an overall policy on tobacco consumption, are discussed.