Last week (Nov. 9), Brad Rodu posted an interesting note on his blog
Tobacco Truth
comparing and contrasting the NHIS smoking prevalence findings with those of another federal government survey (National Survey on Drug Use and Health, or NSDUH), with the latter survey finding higher smoking rates during the past several years.
It is very likely that NHIS and BRFSS findings on smoking prevalence (and perhaps even NSDUH) are lower than is actually the case, as the estimated numbers of smokers multiplied by the estimated average number of cigarettes smoked only account for about 70%-80% of the 17 billion cigarettes sold in the US in 2008.