Steve Becker and Jay Drummond had everything to lose.

Arriving at Fort Belvoir in March for a three-month engineering course, the two young Army lieutenants seemed headed for bright careers in the military.

Becker was an ex-jock, an ROTC graduate who had racked up a string of commendations. "Academically sharp and physically tough," wrote an admiring commander after Becker completed Airborne School in 1981. "One of the most outstanding cadet officers in the corps."

Drummond was a graduate of Texas A&M University whose performance at Fort Belvoir propelled him to the top 20 percent of his class; in a 40-man platoon, his peers rated his leadership skills third from the top.

Then one day in April, all of that ceased to matter.

Subjected to random urinalysis as part of the Army's routine drug testing program, Becker and Drummond, both 25 years old, tested positive for high levels of cocaine. Both men say they have never used the drug. The Army says the tests do not lie, and it stands by the results -- a permanent black mark that all but guarantees the end of an officer's career.

Neither man can prove that he was falsely accused, that his urine sample was mishandled or tampered with. Subsequent tests on both samples have confirmed the presence of cocaine. Nevertheless, the case has raised some intriguing questions.

One has to do with Fort Belvoir's drug screening lab, where the sergeant in charge admitted to having sex with a young woman in the facility on two occasions in the spring, according to documents obtained by The Washington Post. The woman, a private at the base, later alleged that the sergeant had offered to switch her urine sample when she expressed fears of testing positive.

On another level, the case suggests the kinds of problems -- ethical, legal and practical -- that critics contend are an inevitable part of any large-scale drug testing program.

"Everyone who has talked about this realizes that it's going to create some real administrative and bureaucratic problems," said Rep. Steny Hoyer (D-Md.), who has opposed Reagan administration plans to expand drug testing to federal workers. "Clearly with the kind of tests that are involved, one of the concerns that everyone has raised is the possibility of mistakes. This is an example of that possibility."

In some ways, the situation recalls the widely publicized case of former Naval Academy midshipman Jeffrey M. Bellistri, who in 1986 was expelled from the academy after being accused of cocaine use on the basis of a random urine test. Bellistri steadfastly maintained his innocence and eventually was reinstated by Navy Secretary John F. Lehman Jr., who said there was insufficient evidence against him.

In a written statement, the Army defended drug testing procedures at Fort Belvoir and the handling of the Becker and Drummond cases. "There is no indication that there has ever been any problem with the accuracy of the Fort Belvoir lab's test results," the statement said.

Becker and Drummond have protested the drug tests with extraordinary vigor. They have appealed to the highest levels of the Army, alerted -- in Becker's case -- a home town member of Congress, and, finally, told their story to The Washington Post.

"I'm getting hung for nothing," Becker said.

The Army tests roughly 1.2 million enlisted men and officers each year, of whom a little more than 2 percent show evidence of illegal drugs, according to Lt. Col. Douglas Lamar Allen, chief of alcohol and drug policy for the Army. While positive test results for officers are "very, very unusual," Allen said, "it does happen."

The drug testing lab at Fort Belvoir handled 5,757 urine samples in the first nine months of this year, according to the public information office there. Evidence of illegal drug use turned up in 44 of those samples, three of them from officers.

Becker and Drummond insist that they did not expect to be among them.

A former high school football player, Becker had graduated with honors from a private, two-year military academy in Missouri. After finishing college in California, he had accepted a commission in the Army Corps of Engineers.

Drummond was a petroleum engineer by training but opted for the Army after he was unable to find a job in the sagging oil industry. He liked the Army, he said, and was looking forward to his next assignment, as a platoon leader with a combat engineering group in Hawaii.

In March, Becker, Drummond and about 120 others arrived at Fort Belvoir in Fairfax County, home of the Army Corps of Engineers, to take a three-month course. Subjects included demolition, bridge building and other engineering skills.

Becker and Drummond were assigned to the same platoon, which consisted of about 40 people. The two lieutenants said they became casual friends.

Each had taken and passed random drug tests before, they said; in any event, a letter sent to them before their arrival included this explicit warning: "You will participate in urinalysis testing while in the Basic Course to detect illegal drug use."

Given that knowledge, "you'd have to be an idiot" to use drugs, Becker said.

On the afternoon of April 28, a Tuesday, senior officers ordered the members of the platoon out of a classroom and into a hallway adjacent to a bathroom.

They entered the bathroom in small groups, urinated into bottles, and left the samples with two platoon "trainers" who sat at a table by the door.

The samples were kept in a locked room overnight, according to the Army, then screened by the biochemical testing center at Fort Belvoir. A more sophisticated lab at Fort Meade confirmed the positive samples -- in this case, Becker's and Drummond's.

Army records indicate that seven people handled Becker's and Drummond's samples between Fort Belvoir and Fort Meade; the two men suggest that their samples could have been switched or confused at any point along the way.

"The urinalysis test is a faceless accuser, where any one of the . . . people who handled the urine could have allowed a switch or adulteration," Drummond wrote in a memo to the base commander. "There is no way for me to identify who it was."

The Army says that Becker and Drummond handed their samples directly to the trainers, answering "yes" when asked if it was their urine. "The donor then initialed the sample bottle label, the trainer counter-initialed the label and a tamper-proof tape was placed over the lid of the bottle," the statement said.

Becker and Drummond did not find out that their samples had tested positive for a month.

At the end of May, senior officers ordered all the trainees with private cars to drive to a parking lot, where dogs sniffed the cars for drugs without finding anything, according to Becker and Drummond. Then the company commander summoned the two officers to his office and delivered the bad news.

"I asked, 'How can this be? Does this happen all the time?' " Drummond recalled. "And he said 'no.' "

The captains assigned to the platoon recommended that the two men be declared "nongraduates" of the engineering course. "There is no place in the Army for an officer who cannot say 'no' to drugs," one wrote.

The two men appealed the decisions to the acting base commander, Col. Robert Hardiman. Assigned an Army lawyer, they began gathering evidence to prove their innocence.

They volunteered to submit to polygraph tests by the Army's Criminal Investigation Division. Drummond flunked his, and Becker registered an "inconclusive" reading, according to Army records. They hired a civilian polygraph examiner, who said they were telling the truth.

A drug counselor at the Fort Belvoir Counseling Center submitted a statement on their behalf saying that he found no evidence to support "the theory that either client used cocaine."

"Both clients are at the top of their class, have excellent ratings and were considered by their instructors to be excellent commander material," the counselor said. He added, "The sample was taken after class on a Tuesday. One wonders how anyone could function in a classroom setting with that much cocaine in one's system."

Their lawyer presented Hardiman with evidence of improprieties at the Fort Belvoir lab. The sergeant who ran the lab said in a sworn statement that he had sex there with an Army private in early April and again about a month later, which led to his reassignment.

In a memo to Drummond, Hardiman dismissed the significance of the sexual liaisons, noting that they occurred after the lab had closed for the day.

The sergeant's "off-duty conduct does not make the chain of custody a 'joke' as you allege," Hardiman said. "You have presented no actual evidence that this conduct in any way affected the processing of your urine sample."

An Army spokeswoman said investigators found "insufficient evidence" to support the private's subsequent claim that the sergeant had offered to switch her drug test.

Dr. William Manders, a retired Air Force colonel and toxicologist who was involved in the military's early drug testing program and is critical of the way it is done today, said the behavior of the laboratory director should not be entirely discounted.

"Individuals who do testing should be beyond reproach," he said. "They should never have any problems with those individuals."

The sergeant declined to comment.

Manders said he thinks the Army could eliminate many challenges to its drug tests if it required two samples instead of one. "When you're dealing with human beings, anything could happen," he said. "There's just no doubt that a second sample in an hour after the first one would protect that individual beyond any doubt."

According to Lt. Col. Allen, however, "the chain of custody is so stringent that the likelihood of {a mistake} is almost nonexistent."

In the aftermath of several court decisions backing military drug testing, Allen said, lawyers for accused drug users have chosen to find fault with the collection process instead.

"We have so many checks and balances," he said. "We're really confident about what happens." Allen dismissed the idea of taking more than one sample. "I have a hard time understanding what two samples gives that one doesn't," he said. "Our procedures are ironclad, so why not three? Why not four? Why not six?"

In any event, the men failed to develop specific proof that they had been falsely accused. In their hearing on July 7, Becker recalled, "Col. Hardiman asked me one question. He said, 'Did you pee in that bottle?' I said, 'Yes.' And he said, 'Well, I think you're guilty.' "

Hardiman declined through a spokeswoman to be interviewed. The Army said Hardiman was "convinced beyond a reasonable doubt by the facts presented at the hearing" that the two men had used cocaine.

Drummond and Becker appealed Hardiman's decision to Gen. Maxwell R. Thurman, commander of the Training and Doctrine Command at Fort Monroe, Va. After reviewing the case, Thurman opted not to overrule Hardiman's decision.

In the meantime, Becker and Drummond have left Fort Belvoir. Becker said he has been told not to expect any more promotions if he stays in the Army, and he has returned to his parents' home in California.

He is awaiting a final decision on his status.

Drummond accepted an honorable discharge and is looking for work in Florida. He applied for a position in the reserves, he said, but he was told his drug record ruled him out.