During and arbitration hearing this past June, Landis vigorously denied using any performance-enhancing drugs, reports CBS News Radio's Steve Futterman.
But the panel has now ruled against Landis. This will also strip Landis of his 2006 Tour de France title, adds Futterman.
The decision leaves Landis with only one more outlet to possibly salvage his title, an appeal to the Court of Arbitration for Sport.
If Landis doesn't appeal, he'll be the first person in the 105-year history of the race to lose the title because of a doping offense.
According to documents obtained by AP, the vote was 2-1 to uphold the results, with lead arbitrator Patrice Brunet and Richard McLaren in the majority and Christopher Campbell dissenting.
"Today's ruling is a victory for all clean athletes and everyone who values fair and honest competition," U.S. Anti-Doping Agency CEO Travis Tygart said.
It's a devastating loss for Landis, who steadfastly has insisted that cheating went against everything he was about and said he was merely a pawn in the anti-doping system's all-consuming effort to find cheaters and keep money flowing to its labs and agencies.
Landis didn't hide from the scrutiny - invited it, in fact - and now has been found guilty by the closest thing to a fair trial any accused athlete will get. This comes after the initial positive test, then a guilty finding by a USADA panel, then the long lead-up to the arbitration hearing, and now finally, this decision.
Landis has a month to file his appeal. He is still weighing his legal options, according to a statement released by his legal team.
"This ruling is a blow to athletes and cyclists everywhere," Landis said. "For the Panel to find in favor of USADA when, with respect to so many issues, USADA did not manage to prove even the most basic parts of their case shows that this system is fundamentally flawed. I am innocent, and we proved I am innocent."
Despite the result, it's hard to see this as a total victory for USADA, which prosecuted the case. This was a costly affair for the agency, and it exposed flaws in the system.
In its 84-page decision, the majority found the initial screening test to measure Landis' testosterone levels - the testosterone-to-epitestosterone test - was not done according to World Anti-Doping Agency rules.
But the more precise and expensive carbon-isotope ratio analysis (IRMS), performed after a positive T-E test is recorded, was accurate, the arbitrators said, meaning "an anti-doping rule violation is established."
"As has been held in several cases, even where the T-E ratio has been held to be unreliable ... the IRMS analysis may still be applied," the majority wrote. "It has also been held that the IRMS analysis may stand alone as the basis" of a positive test for steroids.
The decision comes more than a year after Landis' stunning comeback in Stage 17 of the 2006 Tour, one that many people said couldn't be done without some kind of outside help.
"It's not a great surprise considering how events have evolved," Pat McQuaid, leader of cycling's world governing body, told the AP by telephone. "He got a highly qualified legal team who tried to baffle everybody with science and public relations. And in the end the facts stood up."
McQuaid said Spanish rider Oscar Pereiro, who finished second to Landis in the 2006 Tour, would be declared the Tour de France winner, as called for by UCI rules.