Cheating in sport is as old as sport itself. The athletes of ancient Greece used potions to fortify themselves before a contest, and their modern counterparts have everything from anabolic steroids and growth hormones to doses of extra red blood cells with which to invigorate their bodies. These days, however, such stimulants are frowned on, and those athletes must therefore run the gauntlet of organizations such as the World Anti-Doping Agency (WADA), which would rather that athletes competed without resorting to them. The agencies have had remarkable success. Testing for anabolic steroids (in other words, artificial testosterone) was introduced in the 1970s, and the incidence of cheating seems to have fallen dramatically as a result. The tests, however, are not foolproof. And a study just published in the Journal of Clinical Endocrinology & Metabolism by Jenny Jakobsson Schulze and her colleagues at the Karolinska Institute in Sweden suggests that an individual's genetic makeup could confound them in two different ways. One genotype, to use the jargon, may allow athletes who use anabolic steroids to escape detection altogether. Another may actually be convicting the innocent. The test usually employed for testosterone abuse relies on measuring the ratio of two chemicals found in the urine: testosterone glucuronide (TG) and epitestosterone glucuronide (EG). The former is produced when testosterone is broken down, while the latter is unrelated to testosterone metabolism, and can thus serve as a reference point for the test. Any ratio above four of the former to one of the latter is, according to official Olympic policy, considered suspicious and leads to more tests. However, the production of TG is controlled by an enzyme that is, in mm, encoded by a gene called UGT2B17. This gene comes in two varieties, one of which has a part missing and therefore does not work properly. A person may thus have none, one or two working copies of UGT2B17, since he inherits one copy from each parent. Dr. Schulze guessed that different numbers of working copies would produce different test results. She therefore gave healthy male volunteers whose genes had been examined a single 360mg shot of testosterone (the standard dose for legitimate medical use) and checked their urine to see whether the shot could be detected. The result was remarkable. Nearly half of the men who carded no functional copies of UGT2B17 would have gone undetected in the standard doping test. By contrast, 14% of those with two functional copies of the gene were over the detection threshold before they had even received an injection. The researchers estimate this would give a false-positive testing rate of 9% in a random population of young men. Dr. Schulze also says there is substantial ethnic variation in UGT2B17 genotypes. Twothirds of Asians have no functional copies of the gene (which means they have a naturally low ratio of TG to EG), compared with under a tenth of Caucasians—something the anti-doping bodies may wish to take into account. In the meantime, Dr. Schulze's study does seem to offer innocents a way of defending themselves. Athletes Raveling to Beijing for the Olympic games may be wise to travel armed not only with courage and the 'spirit of Olympianism', but also with a copy of their genetic profile, just in case. According to the passage, cheating in sport