Cigarettes have changed dramatically over the last 50 years as have the emissions of toxic smoke constituents, as measured by standard Federal Trade Commission (FTC) machine-smoking method. Whereas in 1967 only 2% of cigarettes on the US market delivered 15 mg or less tar, in 2003 84.9% of brands fell into this category. There is a current scientific consensus that yields per cigarette of tar, nicotine, and other smoke constituents measured using the FTC protocol do not provide valid estimates of human exposure or of relative human exposure when smoking different brands of cigarettes. Epidemiological and other evidence does not indicate a benefit to public health from changes in cigarette design and manufacturing over the last 50 years. Currently, the tobacco industry is promoting some new products with claims that they are less harmful because these products purportedly deliver lower amounts of toxic, carcinogenic, and /or addictive agents to the user, compared with conventional products. However, to date, the scientific evidence is insufficient to evaluate whether these new products actually reduce the users' exposure or risk for tobacco related diseases. The key research question that needs to be answered to avoid mistakes of the past is: “Do potential reduced-exposure tobacco products provide a truly less harmful alternative to conventional tobacco products, both on the individual and population level?”