Information on the causes of death obtained from death certificates in Thailand is incomplete and inaccurate. Therefore, mortality statistics from death registrations (DR) remains unreliable. Accurate mortality statistics is essential for national policies on intervention and care and resource allocation. Verbal Autopsy (VA) is a more reliable source for cause of deaths than the DR. In this study, the classification of lung cancer deaths in Thailand from 1996 to 2009 was investigated based on a logistic regression model of lung cancer deaths with demographic and medical factors from the 2005 VA data. The estimated proportions of lung cancer deaths from the model were applied to the DR data. The goodness of fit of the model was assessed using the ROC curve. The resulting estimates of lung cancer deaths were higher than those reported with inflation factors 1.54 for males and 1.44 for females. Meanwhile, misclassified cases were reported mainly as other cancer types. There is no evidence of regional variation for lung cancer. The methods enable health professionals to estimate specific cause of deaths in countries where low quality of causes of death in the DR database and reliable data such as the VA data is available. The findings provide useful information on death statistics for policy interventions related to lung cancer prevention and treatment.