Questions of possible use of a banned steroid by Tour de France champion Floyd Landis were raised because of a urine test that spots elevated levels of performance-enhancing testosterone.

The test detects both testosterone and a related steroid called epitestosterone, which is not performance-enhancing. Both are produced by the body and are also made in synthetic form.

Landis's Phonak team said his urine sample showed "an unusual level of testosterone/epitestosterone" when he was tested after his amazing come-from-behind performance in the 17th stage of the race on July 20. The usual ratio of testosterone to epitestosterone is about 1 to 1 or 2 to 1, said Gary Wadler, a physician and member of the World Anti-Doping Agency.

Suspicions of improper steroid use arise when the ratio climbs above four parts testosterone over one part epitestosterone, Wadler said. Officials have not said what ratio Landis's test showed.

Andrew Pipe, a physician and medical and scientific adviser to the Canadian Centre for Ethics in Sports in Ottawa, says that synthetic testosterone is normally injected, but taking it in the middle of an athletic competition would have little effect in boosting performance.

"Anabolic steroids, of which testosterone is the granddaddy, can have a central nervous system effect," he said. "But anabolic steroids largely work by increasing the capacity for training and increasing the bulk and tolerance of muscles. That isn't going to happen in a few hours.

"The effect of the testosterone is not going to be experienced unless there's a very significant training endeavor associated with it as well."

Pipe cautioned that the initial uproar over the high levels of testosterone detected in Landis's system may prove to be premature, depending on the outcome of additional testing that will have to be carried out before a definitive judgment is made. Taken by itself, he said, an elevated testosterone finding in the rider's A urine sample is enough to raise suspicions, but it does not automatically implicate the athlete as a doping cheat.

"I think it's very important that people take a deep breath and understand the implications and significance of what's being reported," Pipe said last night.He said that some men have naturally high levels of testosterone. For this reason, Pipe said it is imperative that the levels found in Landis during the Tour de France be compared with testosterone levels that have been detected in drug tests that have likely been carried out in the past on all the riders, not just Landis.

If no such comparable data is available, Pipe said, the rider will have to be tested again in the future. For this reason, Pipe said he was surprised that the findings from Landis's A sample have been made public. "The last thing I would want is for the suspected athlete to know that we're on to the fact that he or she may have an unusually high TE ratio and that we may have to administer further tests."