Background. While an association between cigarette smoking and depression has been established in Anglo populations, replication of tobacco–depression associations in countries where smoking is growing may provide important new insights. The objectives of this study were to estimate the association of depressive symptomatology with tobacco smoking, number of cigarettes smoked daily, and smoking cessation in a representative sample of the Mexican population.

Method. The data come from the Third National Addictions Survey (1998) conducted by the Mexican Ministry of Health, representative of Mexico's civilian population residing in cities and towns with 2500+ inhabitants, aged 18–64. Part of a multi-stage, stratified, probability sample, 1935 men and women answered a version of the survey that also included the CES-D depression scale. Analyses addressed the survey's complex design and controlled for income and educational level.

Results. Among women only, current smokers had twice the odds of elevated depressive symptomatology than never smokers (OR 2·1, 95% CI 1·3–3·5, p=0·002). For men, only those smoking a pack or more a day had greater odds of depressive symptomatology (OR 5·9, 95% CI 1·6–21·9, p=0·008). Overall, former smokers who ceased smoking within 6 months had lower odds of depressive symptomatology than current smokers (OR 0·4, 95% CI 0·1–1·0, p=0·042).

Conclusions. These findings add to the accumulating evidence for the association between smoking and depression in different cultures and populations.