The literature on the effects of public smoking bans on smoking presents conflicting results and there is limited evidence on their impact on active smoking. We evaluate the impact of smoking bans on active smoking by developing a theoretical model that predicts the likely behavioural effects of the policy. Using data from the British Household Panel Survey (BHPS) we exploit the differential timing of the introduction of the Scottish and English public smoking bans. Our empirical models provide evidence consistent with the prediction of the theoretical model that bans will tend to have a limited effect on smoking prevalence, at least in the short run, and a rather limited impact on the total level of smoking. Also, while we identify significant differences in trends of smoking consumption by population sub-groups throughout the entire survey period, we find insufficient evidence to conclude that the bans affected the level of cigarette consumption. These results challenge those found in the public health literature but are in line with the most recent strand of economic literature indicating that there is no firm evidence on the effects of smoking bans on smoking.