Using lung cancer mortality as an indirect marker for smoking risk, Ezzati and Lopez expanded the Peto-Lopez method to estimate global mortality due to smoking in 2000. The American Cancer Society Prevention Study and the work of Liu et al in China were used to calculate relative risks and these were corrected for confounding variables.

They estimated 12% of total global deaths were attributable to smoking (4.83 million). Roughly equal numbers of people died of smoking related disease in the industrialised and developing worlds (2.43 and 2.41 million, respectively); men accounted for 84% of these deaths in the developing world and 74% in the industrialised world. 2.69 million people died aged 30–69 years, resulting in a substantial loss of productive years. The leading causes of death from smoking were cardiovascular disease (1.69 million), chronic obstructive pulmonary disease (0.97 million), and lung cancer (0.85 million).

The percentage of deaths due to smoking is currently higher in the industrialised world (19%) than in the developing world (9%), but this is likely to change as the rate of smoking increases in the developing world. Initiatives are desperately needed to reduce the rate of smoking in men and to prevent increases in women.