The Linear No Threshold Model
I'm sure that my fellow technically-oriented reactionary crackpots have heard of the report from The National Academy of Sciences purporting to show there is no safe dose of radiation. I'm more than a bit dubious.
First, any organization that puts out press releases should be suspected of selecting data. Given that the effect they predict is barely noticeable (the variation in the expected effect is only slightly less than the effect), it's possible that the minute effect didn't take much selection.
Second, they based most of their conclusions on studies of survivors of Hiroshima and Nagasaki. In other words, this was based on the effects of a single dose per patient. Studies on the effects of chronic doses (for example, studies of the residents of a radioactive building or the residents of high-radon areas) indicate that a dose of 50 mSv per year can lower mortality rates. (The atomic-bomb survivor studies mentioned earlier indicate that single doses of 100 mSv can raise mortality rates.)
It's worth noting that the radiation people are worried about (from nuclear waste or “dirty” bomb) is a matter of chronic doses. (On the other hand, the report from The National Academy of Sciences might possibly have some relevance to judging appropriate x-ray doses.)