Background:
Illicit trade in cigarettes poses a threaten to tobacco control. Reliable measures of the extent of this problem are scarce, however. The objective of this study is to use tobacco surveillance data to measure illicit cigarette consumption in Mexico.

Methods:
Data comes from the GATS Mexico 2009 and 2015. In particular, self-reported information of brands is employed to identify illicit cigarettes. Since the GATS 2015 collected images from packs in smokers' personal possession, we also use this information to validate self-reported data.

Results:
Self-reported information indicates that illicit cigarette consumption increased from 0.5% to 2.7% between 2009 and 2015. For the subsample of smokers with both self-reported and cigarette pack information, the share of illicit cigarette consumption is lower if calculated with the former, but the difference is not statistically significant. Additionally, if we use cigarette pack information to adjust the total estimate for 2015, this increases by just 1.1 percentage point. Only 25% of the smokers had a pack at interview, however. This is related to marked differences in consumption patterns: being a daily smoker and having bought a pack at last purchase, as opposed to being occasional smokers and having bought a single cigarette, increases the odds of having a pack at interview. These characteristics are also associated with the likelihood of consuming illicit cigarettes.

Conclusions:
Despite important increases in tobacco taxes during 2009-2015 in Mexico, surveillance data indicate that illicit cigarette consumption increased by only 2 percentage points and is currently below 5%. This figure is much lower than the estimates of the tobacco industry, which is consistent with the findings of previous academic studies for other countries. Self-reported information was rather accurate for more heavy smokers, but future research should assess whether that is also the case for occasional smokers, who are more likely to buy single cigarettes.