Just two days from the start of the Summer Olympics — and the thousands of drug tests that will be performed during the Games — a researcher at the University of Texas’s M.D. Anderson Cancer Center has come bursting out of the starting blocks with a controversial claim: The drug-testing practices used on athletes are basically worthless.

As a result, Donald A. Berry writes in the new issue of Nature (subscription required), “Cheaters evade detection, innocents are falsely accused, and sport is ultimately suffering.”

One problem, Mr. Berry argues, is that the tests may never have been validated. That is a test itself may not have been tested on clean samples and on those from people taking forbidden drugs, to make sure it can distinguish between them. With such a test, it is impossible to know how many dopers are incorrectly cleared, or how many clean athletes are falsely accused.

“If conventional doping testing were to be submitted to a regulatory agency such as the U.S. Food and Drug Administration to qualify as a diagnostic test for a disease, it would be rejected,” writes Mr. Berry, who is a professor and chair of the department of biostatistics at M.D. Anderson. He urges sports organizations to require laboratories to validate and reveal publicly their testing procedures, including the criteria by which they conclude a sample is tainted.

With so many tests being run at the Olympics, the chance of a false positive goes up. He illustrates the problem with an example of a hypothetical test that correctly determines that 99 out of 100 clean samples are in fact clean. If that test were used on 126 samples, there’s nearly a three-quarters chance that a mistake would occur and a clean athlete would seem to be doping.

Why 126? That’s how many samples were tested in the 2006 Tour de France — an event whose winner, Floyd Landis, was stripped of his title after failing a drug test. —Lila Guterman