The first nuclear bomb was successfully tested in July 1945 in New Mexico. On August 6, 1945, an atomic bomb was dropped on Hiroshima and three days later a second, more powerful bomb was dropped on Nagasaki. About 550 further explosions were carried out by the United States, Britain, the Soviet Union and France before most countries signed the Limited Test Ban Treaty in 1963. China tested its first Bomb the next year.
Dr Basner’s team argue that this activity released isotopes into the environment that do not occur naturally. Tiny traces of these isotopes, caesium-137 and strontium-90, permeated soil and plant life and ended up in all postwar paintings through the natural oils used as binding agents for paints.
Any work of art purporting to be more than 63 years old that registers trace amounts of the two isotopes can therefore be definitively declared a fake, Dr Basner said.

This reminds me of a story out of a National Geographic from the late '60s:

Back in the days when the first computed tomography machines were being developed, it was necessary to obtain steel, to make the frame and base assemblies, that had no radioactive atoms in the mix. The best place to find such steel was in old pre-WW2 warships. I'd have to ask my brother, who runs the current CT and MRI scanners, what is used now.

These traces got into the soil and the air and found their way into new paint but somehow didn't get into paint already on canvas?

I don't follow.

Also, how far did radioactive fallout from the bombs get, and does the concentration vary from location to location? I'd assume that the concentration would be greater closer to Japan, or to a bomb test site, and if that wasn't taken into account, it could throw off the results.

Wasn't this already covered in a movie??

Radiation from nuclear fallout travels globally since vaporized material is flung high into the stratosphere during an explosion.

I learned in school that anyone born post WWII is 4 to ten times more radioactive than their grandparents.
(An Italian supply cache from WWII found in the Sahara in the mid seventies was widely sought after by researchers. Its barrels of drinking water were the only known source of pre-war water uncontaminated by fallout. It was used to establish a baseline for how much back ground radioactivity had increased since the war.)

Oh, and good news for those born from the late 50s to mid sixties; Since nuclear testing was at its height then, your brains are 10-18 percent more radioactive than any other generations!

Because brain cells aren't replaced like other body cells, they retain the carbon 14 that was produced in elevated levels at the time.
(Beware the nuclear brained baby boomers!)

I feel that I should allay fears that my above post might cause. Even an 18 percent increase in background radiation is negligible.
In fact, there's evidence that slight increase actually decreases you risk of cancer!

If you're still nervous, check out this site:
http://home.earthlink.net/~bhoglund/radiation_Facts.html

Also, even if the forgeries were closer or even in Japan at the time of The Bomb, this method would still work. Standard radiometric dating, right? Compare the isotopes currently detected vs their decay products. So something closer to the initial blast/fallout would have both the higher initial isotopes and a corresponding higher 'bi-product' elements. Like comparing 5/10 vs 500/1000. Still get the same number.
Heck, Caesium-137 has a half-life a little over 30 years, and Strontium-90 about 29 years. In theory using this they could even get pretty close on -when- it was forged. Proly not the exact year, but within a 5-10 year estimate.
And I thought I read about this too..years ago, proly in a SciFi book. But the Science is sound. I'm pretty sure someone has used the same thing(A-Bomb radiometrics not the regular radiometrics) for some other issues, but the 1st time I heard it applied to art.

You would have to have a way to determine either the inital amount of Cesium and strontium incorporated or the amount of product formed, in addition to the amount left, not only the amount that is left, in order to determine age.

This theory it was tested/covered on an episode of a police drama, CSI: New York--'Tri-Borough'

http://www.csifiles.com/reviews/miami/tri_borough.shtml

For this method to work, it wouldn't matter how near or far the painting was in relation to Japan or the various test sites. Nor would it matter exactly how old it was, since it isn't designed to date paintings exactly-- only to determine if they were painted before or after 1945. Caesium-137 and strontium-90 do not occur naturally. They are examples of "rare earths": elements created only by nuclear fission. Since there was NO strontium-90 nor caesium-137 in the world before 1945, if there is ANY embedded in a painting (it doesn't really matter how much), the painting was made (or at least altered) no earlier than 1945. If the painting has NONE of these elements, it was likely made before 1945.

Any fallout adhering to an old painting would likely be in the form of surface dust, which could be carefully cleaned off before the painting was tested.

You could get old paint by taking paint from anything painted in the period you want to imitate, grinding it up and mixing it with old oil (which can also be obtained). Let's say you want to fake a painting from 1750. You find a relatively unvaluable painting from 1750, clean off the canvas, remix some 250-year-old paint, paint exactly the way the masters of that era did, and you're in business.

Old documents have been forged fairly convincingly using the same basic method: Bleach a piece of old paper (to remove whatever was on it before), then make some ink by burning some paper the same age, and do your writing. It stands up to carbon dating very nicely.

Just found this site by accident...Anyhow, to "JoeDaJuggler": The radioactive materials in the air/water/soil get pulled into the plants that produce the linseed/walnut oils that are the major binding agents in traditional mediums.

A paint created before the testing wouldn't be effected and you can't use synthetic oils as they wouldn't have been produced in the periods an artist would be mimicking.

It sounds interesting.But I wonder the method does work.I guess it would be a hard task to detect.