A piece of contrarian science that could have come straight from the protagonist of Thank You For Smoking, The Scientific Scandal of Anti-smoking. The results of the Whitehall study cited in the article are that you don't see much difference in life expectancy between smokers who quit and smokers who just keep on puffing away.
Robin Hanson looks into some of the other studies cited and finds more of the same. If you look at smokers versus non-smokers, there is a big difference in life expectancy, but if you take people who chose to smoke and then convince them not to, you don't see a big improvement. This could indicate the presence of confounding factors that are correlated with smoking. Smoking is obviously bad for you, but just how bad is actually less clear than one might think.
I especially like the point made in the article that Japanese people persist in smoking at high rates, and maintain a high life-expectancy regardless. I also like to use Japanese people to point out the clear absence of mercury poisoning from eating lots of fish.