In 1969, a year after the Air Force had awarded a contract for development of the first-generation Maverick missile, its preferred anti-tank weapon, the President's Science Advisory Committee scolded the Pentagon for rushing to buy arms before finding out whether they would work in battle.

"Most of our previous failures to be prepared for wars we fight would have been thoroughly exposed had an adequate program of testing and evaluation existed," the committee said.

Nevertheless, the Air Force went ahead with production of the Maverick in the early 1970s.

In May, 1980, 11 years after the committee's warning, Robert A. Moore, deputy undersecretary of defense for research and development, had this to say to the House Appropriations defense subcommittee about the first Maverick: "I think that is one case where, in my opinion, R&D research and development failed . . . because we did not test that weapon in a realistic operational environment before we decided to buy.

"That was a mistake, and from that, I think the lesson we should have learned is that we must test those sophisticated weapons in a realistic operational environment . . . ."

The total cost of the first-generation Maverick was $500 million. There are now 20,000 of them stockpiled by the Air Force, and the missile is still the main Air Force anti-tank weapon in Europe.

Gen. John W. Vogt commanded U.S. air forces in Vietnam, where the Maverick failed, and later headed the air forces in Europe, where he insisted on not buying any more. Vogt's assessment of the Maverick: "To have to rely on it as the main anti-tank weapon makes no sense."

How, over the last 13 years, did the Air Force and Pentagon oversee what they acknowledge as a research and development disaster?

The reason for asking that question, say a number of military and civilian weapons experts, is that the Pentagon could repeat its mistake.

The Air Force wants to spend about $5 billion on a second-generation Maverick, a more sophisticated and complicated missile that has tested poorly even in favorable environments that are not close to combat conditions.

The full-production decision yesterday was put off until 1984. But next Tuesday a panel of top Defense Department officials is expected to give the Air Force the go-ahead for the initial purchase of several hundred test Mavericks.

Back in 1969, when the President's Science Advisory Committee rebuked the Pentagon, President Nixon and Defense Secretary Melvin R. Laird commissioned a blue-ribbon panel to take a fresh, year-long look at the operation and structure of the Defense Department. The members said they were appalled by their findings on weapons testing.

"The vast, horrendously expensive weapon systems which now consume so large a part of the budget of the Department of Defense may be our saving or our downfall," said one member, George J. Stigler, professor of American institutions at the University of Chicago. "The great difficulty is that presently we do not know."

The difficulty could be solved by operational testing, which exposes weapon systems to realistic simulations of combat. Yet such testing, the panel found, was "too infrequent, poorly designed and executed, and generally inadequate."

The panel also discovered it was standard practice to let the armed service sponsoring a weapon system evaluate the testing of it.

On Capitol Hill, growing disillusionment with the Pentagon's preference for expensive Buck Rogers weapons has led even pro-military legislators to express open concern.

"Public support for increased defense spending will not be sustained for the long haul if the Pentagon cannot keep the people's confidence that every dollar is spent responsibly and efficiently," Sen. John C. Stennis (D-Miss.) said recently.

As former chairman of the Senate Armed Services Committee, Stennis has followed the development of the Maverick since its earliest incarnation.

At about the same time that both presidential groups criticized the Pentagon for its testing program, Stennis asked the General Accounting Office, the investigative arm of Congress, to study the proliferation of the tactical air-to-ground missiles.

One of these was the first-generation Maverick, an eight-foot-long guided missile carried on aircraft for launching at ground targets, primarily tanks.

By November, 1970, the Air Force had geared up to order into production the original Maverick missile, which uses a television sensor to guide it to its target.

In December, the GAO responded to Stennis' request with a secret report. A sanitized version shows a litany of worries:

The TV Maverick shared "the common weakness of all television-guided weapons to date . . . they can be lured away from the target by nearby objects having stronger contrasts." It "will often require the pilot to identify small, fleeting targets, which are sometimes barely discernible from even short distances ." "Smoke, haze, and camouflage will inhibit it in the battle area."

In addition, the report said: "For the close-air-support mission, most guided missiles to date involve danger to friendly troops . . . Although the Maverick is designed for close support of Army troops, the Army does not plan to participate in the testing and has had little to say in the development of the missile."

Moreover, available analyses "indicated substantial doubt that average user pilots could detect enemy tanks in time to launch the missile." If a pilot must "run in and see" tanks and then back off to fire the missile, his protection from enemy fire "is diluted."

In addition the report said that operational testing and evaluation were "more or less emasculated."

The GAO's conclusion, which it repeated and emphasized, was that the "factual evidence" was too inadequate to "justify going into production." Taking the advice of the House Armed Services Committee, Congress refused to appropriate Maverick procurement money for fiscal 1971, in order to allow time for "more reliable development and test results."

Tests were then conducted at Nellis Air Force Base in Nevada and Holloman Air Force Base in New Mexico. The dark target tanks on the light desert sand provided an ideal visual contrast for the television sensor.

For fiscal 1972, the Air Force requested $74 million for full-scale procurement. House Armed Services approved the request in May, 1971, saying that "test results . . . have been highly satisfactory." In September, Stennis' committee called the same tests "favorable."

The full Congress agreed, and the Air Force ordered production to start.

"It would have been simple to discover the target acquisition and attack problems with inexpensive flight tests prior to making a heavy investment in missile development," said C. E. Myers Jr., who was the Pentagon's director of air warfare from 1973 to 1977.

"A realistic operational assessment of this information should have discouraged the development of a forward-fired, precision-guided anti-tank missile for high-speed aircraft."

A former Pentagon weapons analyst said that "a few test seeker heads could have been attached to airplanes and tried out, and this would have shown that the whole TV . . . thing was a bad idea. The cost would have been only a few million dollars."

Hughes Aircraft, which makes the missile, delivered the first TV Maverick three weeks ahead of schedule in August, 1972, which allowed ample time for its use in battle in Vietnam. The Air Force sent the missiles to Gen. Vogt, then the commander of U.S. air operations in Southeast Asia.

Nineteen months later, the Air Force acclaimed Maverick's success in battle. "We had good luck," Lt. Gen. William J. Evans, deputy chief of staff for research and development, told the Senate Armed Services Committee.

Vogt flatly contradicts this.

Vogt had tried to use the missiles to stop traffic on the Ho Chi Minh trail. The Maverick requires that the pilot of the aircraft carrying the missile take the plane to a higher altitude to have a straight flight path before launching. But Vogt said in an interview recently that the "pop-up" maneuver alerted the enemy ground forces and gave them time to blend with the foliage or the shade. This degraded or blotted out the visual contrast between the target and background that the TV Maverick was designed to sense and home on.

Vogt said that Evans' claim of success was "obviously wrong." In summary, Vogt said, the TV Maverick "wouldn't work . . . . Clearly, it didn't do the job out there."

By 1975, even after this failure in Vietnam combat, Vogt says he heard the Air Force was planning to buy about 6,000 more Mavericks, primarily for Europe. Vogt then was commander-in-chief of U.S. air forces in Europe, and he ordered what he says was the first true operational test in combat-like conditions.

The tests, conducted at Hahn Air Force base in Germany from March 10 to April 30, 1975, showed that the TV Maverick's performance was "very poor," according to Vogt, who declines to be specific.

Other officials who had access to the test results say that of the attempts to acquire target tanks for a "lock-on" of the TV guidance system only 23 percent were successful.

When attempting to get the sensor to hold a lock on a tank at the edge of a tree line or against camouflaged tanks, the success rate was extremely low, under 5 percent. Against moving tanks, according to the officials, the outcome could not have been worse--not a single lock-on was held long enough for a successful firing.

Given these results, Vogt told the Pentagon that the Air Force in Europe had no requirement for more TV Mavericks.

To this day, however, the TV Mavericks are stockpiled as the Air Force's primary anti-tank weapon for Europe.

The Air Force let three years go by after the European test before halting production of the TV Maverick. Even in 1978, however, it was being applauded by Evans' successor, Lt. Gen. Alton D. Slay, at a hearing of the House Armed Services research and development subcommittee.

Rep. Thomas J. Downey (D-N.Y.) told Slay that Vogt "thinks, or used to think, that this weapon was useless."

A touch of indignation marked Slay's response: "Gen. Vogt would be the first man to stand up here and tell you that, compared to any other weapon in the inventory today, the TV Maverick is pretty doggone good, sir."

Recently, Slay's "pretty doggone good" comment was read to Vogt, who retired at the end of his tour of duty in Europe and lives in Annapolis. He laughed. The testimony was read again and Vogt laughed again, adding, "I think Al is putting words in my mouth, quite obviously."

Back in 1970, when the GAO raised its questions about the TV Maverick to illustrate what it called the "unnecessary proliferation" of weapons systems, the GAO said in its report:

"If the weapon falls short of the requirement, a successor design must be developed and produced. There will then be two weapons in inventory when only one is necessary."