Some policy makers justify cigarette taxes by arguing that they actually make smokers better off. This argument has been hard to evaluate because behavioral data, such as that showing reduced cigarette consumption following a tax hike, cannot resolve the issue of whether smokers are made better off by the reduction or not. In this paper, we directly assess the effect of cigarette taxes on well-being, using subjective well-being data. We model the differential impact of excise taxes on those with a propensity to smoke, relative to others, in order to control for omitted correlations between happiness and excise taxation. Using US data on happiness and state-level changes in excise taxes, we find consistent evidence that excise taxes make those who have a propensity to smoke happier. To assess robustness, we repeat the exercise using Canadian data, which has independent information on well-being and also much larger tax changes, and find the exact same pattern. Moreover, these impacts are present for cigarette excise taxes, but not for other excise taxes. These results suggest that the welfare effects of cigarette taxation are far more complex than simple rational economic models might predict.