Nine members of the Army's select White House Guard Company have been placed on company restriction and are being reassigned after unannounced urinalysis tests revealed traces of marijuana.

The test was administered to the 50 members of 1st Platoon, E Company, 3rd Infantry. The Army White House Guard, an elite corps of 200 enlisted men who perform ceremonial duties, is attached to the Old Guard at Fort Myer, the oldest active infantry unit in the Army.

The results of the March 19 tests were received the first of this week, according to Col. Jamie Walton of the public information office of the Military District of Washington.

The offending soldiers, Walton said, are now awaiting reassignment from the 3rd Infantry to other Army units. Disciplinary action in the form of nonjudicial punishment--extra duty, suspension of pay, or reduction in grade--is also pending.

Two of the soldiers, contacted yesterday, declined to be quoted by name. One, a 23-year-old private, said "I don't smoke marijuana. I'm around some people who do, but I don't smoke it myself.

"One morning they told us we were going to take a test. They handed out specimen bottles. I have no idea why they did it," the private said. The nine soldiers, he said, asked to be retested but were refused.

Attempts to reach the soldiers' commanding officer were unsuccessful.

The White House declined comment on the incident through a press spokesman yesterday.

Random urine tests for marijuana use have been administered in the armed forces, at the discretion of individual commanding officers, since the beginning of last year. However, in an effort to crack down even more heavily on drug use, the Army will begin on July 1 processing for discharge all officers, noncommissioned officers and senior enlisted soldiers found to be drug users, and all enlisted soldiers determined to be second-time drug abusers.

"Things are going to get very, very tough. The Army is putting out notice that drug use will not be tolerated," said Margaret Tackley of the Army public affairs office.

Since February 1982, when the Department of Defense began giving the urine tests to U.S. soldiers on a worldwide basis, Army spokesman Walton said, about 10 percent of those in the White House Guard tested showed signs of marijuana use and were immediately reassigned. Walton could not say how many soldiers that represented, except that it was "a small number."

A survey of drug use in the armed forces, to be released soon, shows that 22 percent of enlisted men and women use marijuana at least once a month, down from 37 percent in 1980, the Defense Department said.

"There has been a substantial decrease of drug abuse in the armed forces, and we are ascribing that to the increased use of urinalysis," said John Allen, with the department's office of health promotion.

The urine tests for marijuana, introduced commercially in 1980, are considered accurate up to 14 days after use of the drug, according to researchers at SYLVA Co. of Palo Alto, Calif., the company that pioneered the commercial tests. Current technology, however, can measure neither the quantity used nor whether the usuer was actually high on the drug.

In addition, conflicting claims about accuracy have arisen. The National Institute on Drug Abuse claims a 95 percent or higher accuracy, and the military says it's methods are virtually foolproof. But critics, like the National Organization for the Reform of Marijuana Laws, claim the tests are only 50 percent accurate and that they can also wrongly accuse someone who "passively inhaled" marijuana fumes.

Human error can be a factor as well. In 1982, for instance, the D.C. police tested recruits and found that 39 had recently used marijuana. After the recruits alleged that the bottles containing the urine samples were mislabeled, misplaced and possibly switched, 24 of the recruits were reinstated.