The panel structure of the Survey on Smoking in Canada (1994-95) and novel methods are used to estimate the impact of an important decrease in the levels of taxation of cigarettes occurring in five out of the ten Canadian provinces that intended to eradicate black market sales of cigarettes in the spring of 1994. Given that black market sales have recently increased substantially because of new taxes, a complete and thorough analysis of the 1994 policy is of particular importance for policy makers. We revisit the issue with new econometric methods to address this evaluation problem as well as focus on particular sub-groups in the Canadian population. The large sample permits precise estimation of the impact of the policy by sub-group showing that females, young males, the poorly educated and separated or divorced individuals were particularly sensitive to these dramatic changes in cigarette prices. We also compute under realistic assumptions a price-elasticity for the probability of smoking and a lower bound on the price-elasticity for the quantities of cigarettes smoked.