ON SEPTEMBER 5, seven-time Tour de France champion Lance Armstrong made a remarkable announcement: He said that he was thinking of ending his retirement, at that point a mere six weeks old, and rejoining his Discovery Channel team for a run at another title. Although Armstrong, 34, had previously said "only an absolute miracle" would get him to race again, his rationale for a return in 2006, as published in the Austin American-Statesman, was loud and clear: "I'm thinking it's the best way to piss [the French] off."

Armstrong's motivation for wanting to anger the French was equally clear: It would be payback for the publication, on August 23, of a controversial story in L'Équipe, a Paris-based sports daily. Under the headline "Le Mensonge Armstrong" ("The Armstrong Lie"), investigative journalist Damien Ressiot, 41, alleged that he had proof that Armstrong had used a banned performance-enhancing drug during his first Tour victory, in 1999. "The facts are indisputable," he wrote.

L'Équipe's proof, Ressiot reported, had been obtained by comparing the anonymous results of urine testing conducted by France's top anti-drug lab, the Laboratoire National de Dépistage du Dopage (LNDD), with identifying numbers from six separate doping control reports naming Armstrong—one of which could be identified as having come from Union Cycliste Internationale (UCI), pro cycling's Aigle, Switzerland–based governing body, which sanctions races, licenses riders,

Walsh remains unrepentant about the allegations put forward in his book. "I find it hard to believe a person could read our book with an open mind about Lance," he says, "and come away still believing in him."

and runs its own anti-doping procedures. The doping lab had evaluated the urine samples, taken from Tour riders in 1999 by the UCI and frozen for five years, in the course of a 2004–05 laboratory research project to refine a new testing method. The tests, according to the L'Équipe article, showed that some of the urine samples contained traces of erythropoetin (EPO), a red-blood-cell booster first used in the late eighties to increase endurance. Six of the samples that tested positive, Ressiot charged, were Armstrong's.

No official from the LNDD or any other anti-doping body had identified Armstrong; the research was not part of any enforcement protocol. Consistent with standard procedure, the lab had no knowledge of whose samples it was testing, and the samples were identified only by numbers. Ressiot claimed that he had the six doping control records—including the UCI file—that linked Armstrong to the identifying numbers on the samples, and that the LNDD had used the same numbers to identify the positive samples from the LNDD test. An unnamed source or sources provided Ressiot with the evidence he used to make the link.

Ressiot's talent for collecting this type of information has prompted a separate investigation of his methods: On October 13, French authorities announced they were looking into his use of confidential police interviews in an April 9, 2004, L'Équipe story about alleged doping by members of the French Cofidis cycling team. But he makes no apologies about his reporting style, and in an interview with Bicycling magazine, he has characterized his Armstrong story as "black-and-white evidence."

The story went off like a bomb. Some observers said the charge was a grave one and that the case for Armstrong's guilt was compelling. Others raised a flurry of criticism attacking the ethics and legality of leaking confidential lab results, the motives of the journalists and sources involved, and the accuracy of the test in question. As happens whenever Armstrong is accused of anything, many of the reactions on both sides were characterized by greater certainty than the available evidence seemed to warrant. Jean-Marie Leblanc, the Paris-based director of the Tour, called the story "meticulous," adding later that "we were all fooled" by Armstrong's declarations that he had raced clean. Dick Pound, chairman of the World Anti-Doping Agency (WADA), an independent Montreal-based operation established by the International Olympic Committee to set testing policies for most of the world's sports federations, was more circumspect. He would not commit to a direct accusation against Armstrong but declared that the L'Équipe report showed there was "a very high probability of performance-enhancing-drug activity" at the 1999 Tour.

In Armstrong's defense, Gerard Bisceglia, CEO of USA Cycling, American cycling's Colorado Springs, Colorado–based governing body, called the paper's story "preposterous," adding that "this kind of years-ago testing of a single sample with new technology is completely without credibility."

Bisceglia's reference to "a single sample" pointed to a crucial fact: It's an ironclad rule of current anti-doping regimens under the auspices of WADA that a sanction against an athlete is allowable only when two samples from a single urine or blood specimen—known as A and B samples—both confirm a result. But the test in question had been conducted entirely with B samples; the A samples from the 1999 Tour had long since been destroyed.

In fact, the lab's real purpose wasn't to determine evidence of doping in a past event but to refine the test itself. The standard EPO urine test used today works via electrophoresis, which produces a kind of electrically charged photograph that distinguishes between isoforms—functionally similar proteins that have different genetic codes. On the snapshot, the body's naturally occurring EPO looks different from artificial EPO. To test positive for doping, at least 80 percent of the isoforms have to show markers that appear consistent with manufactured EPO, and the sample must satisfy visual and mathematical analysis. It was hoped that a refined test could provide a qualitative positive—one that would be based not on a threshold but on how, exactly, manufactured EPO isoforms look.

The standard test, first developed in 2000 by LNDD, is by no means lock-tight, since its subjective nature raises the possibility of false positives. In fact, two weeks before the L'Équipe story was published, elite triathlete Rutger Beke got a 2004 positive EPO reading thrown out on appeal by the Flemish Disciplinary Commission when he demonstrated hat he had naturally produced proteins that could trigger a positive result.

L'Équipe's story also raised questions about athletes' right to be tested and judged by the rules, not by a newspaper sting that wouldn't hold up under the standards of the official anti-doping protocol. Another criticism of the exposé was that Armstrong was the only cyclist to be named, despite evidence of other riders testing positive. (Two and a half weeks after the L'Équipe story broke, the French newspaper Journal du Dimanche published a story naming three other cyclists whose samples allegedly tested positive during the LNDD research project.)

The UCI, which has jurisdiction in the matter because it was in charge of testing at that time, opened an investigation, but its attention quickly diverted from the doping question to the sources of the leaked lab report and Armstrong's individual doping control forms. Their actions led to an international war of words between then UCI president Hein Verbruggen and WADA's Pound over each agency's role in the scandal.

"Pound loses his mind when the press gets near," said Verbruggen, responding to Pound's claim that Verbruggen himself was the source of the leak. To sort the matter out, on October 6 the UCI announced that it had hired a new independent investigator, Emile Vrijman, a Dutch lawyer and former head of Holland's anti-doping agency, to look into the test results and the leaks surrounding the L'Équipe affair.

Ultimately, the most important reaction came from Armstrong himself. The night before the story broke, he published a denial on his Web site, LanceArmstrong.com. "I will simply re-state what I have said many times," he wrote. "I have never taken performance-enhancing drugs." He subsequently told the Associated Press, "There's a setup here. . . . I absolutely do not trust that laboratory." On August 25, he appeared on CNN's Larry King Live and speculated that the only way he could have tested positive was if his urine sample had been tampered with. "What was manipulated was the urine," he told King and co-host Bob Costas. "What was put in the urine? Who was there?"

In the end, Armstrong withdrew his threat to return to racing. During a September 15 conference call with the media from Austin, he ruled out a return to competition. "There is no way I would get a fair shake," he declared, "either on the roadside, in the doping control, or in the lab, or in the hotel, or in the food."

During the Larry King Live interview, Costas asked Armstrong whether he was considering a lawsuit against either L'Équipe or its sources. He had thought about it, Armstrong said, but had decided against it. "At the end of the day, when you sue somebody, it just keeps a bad story alive forever," he argued. "It gives them the opportunity to say, ‘Oh, we found this. Oh, we did that.' It gives them more credit than they deserve."

As the noise from all sides died down to a low yet consistent rumble, cycling fans around the world were left wondering what had happened. Damien Ressiot's L'Équipe story, it seemed, produced more questions than answers.

Meanwhile, however, with little scrutiny from the press or the public, a legal drama is unfolding that could finally provide a definitive answer to the question that has dogged Armstrong for seven years: Did he rely on doping to win the Tour de France—or has he been the victim of a callous and vicious witch hunt?