OBJECTIVES: Smoking is a leading cause of early dead and morbidity in the Western World. The purpose of this study is to evaluate how tobacco sales and tobacco survey data correlates with lung cancer incidence in the Danish population. It is estimated that up to 86% of lung cancer cases in developed countries are smoking related. METHODS: Lung cancer incidence data from 1943-2009 are from Nordcan (www.ancr.nu). Sale of tobacco (cigarettes/inhabitant) 1920-2010. Smoking habit surveys from 1953-2010, annually from 1969. Lung cancer incidence is age standardized to the Nordic population (ASR(N)), and is in rate per 100.000/year. Correlations are analyzed with Spearman's rho with SPSS18. RESULTS: The strongest correlation (spearman’s rho = 0.92, p<0.0001) is found between sale of cigarettes and incidence of lung cancer with a lag time of 24 years. The correlation between lung cancer and the proportion of the population that smokes is well correlated for men (0.8, p<0.0001, lag time=20 years). Female smokers and lung cancer are with a lag time of 5-26 years negatively correlated, but correlates positively when the lag time is more than 27 years, the best correlation being 0.732 (p=0.039, lag time=35 years). CONCLUSIONS: The correlation between lung cancer incidence and the sale of cigarettes is better than for the proportion of smokers. This might be because sale gives a better estimation of the overall exposure in a form of population “pack years”. The negative correlation between the proportion of female smokers and lung cancer, and the change to a positive correlation when a longer lag time is applied can be either a true finding that might be explained by longer development time in females. Or it could be a result of changes in the accuracy of the proportion, or a result of changes in the age pattern.