Introduction:
France has an erratic anti-smoking policy. Effectives measures follow phases of stagnation or even regression. Until recently, taxation policies were decided only in the short term. These variations give the opportunity to analyze the effect of the price increase on cigarette sales.

Methods:
From 2000 to 2018 we study change in yearly sales of cigarettes in metropolitan France to (source OFDT) to the price of the top sale cigarette pack (year n/ year n-1). In the case of an increase during the year, a monthly weighting is carried out. The results are expressed as the elasticity. If a price increases by 10% and sales fall by 6% the elasticity will be 0.6, it will also be 0.6 if a 5% increase price leads a 3% drop in sales).

Results:
From 2000-2018, the 3 price increases >10% caused a sharp tobacco sales decline (0.9 elasticity). The 2 increases of 6-9% were almost as effective (elasticity 0.84). The 2 increases of 5% price resulted in a lower response (0.47 elasticity). The 5 small price increases <4.5% provide minimal response (0.16 elasticity). Increase of cigarette sales has been observed during the 4 years without price increase.

Conclusions:
An above 6% step price (taxes) increase is needing to have significant public health effect. A 10% step increase is optimal Therefore any increase of <5% increase revenues of states without health benefits. Step of increases must be > 6% or better 10% to lower tobacco consumption without reducing the resources of the states.