When William P. Clark Jr. replaced Richard V. Allen as President Reagan's national security affairs adviser in 1982, the 40 professional staffers who advise the adviser noticed a remarkable change.

Assistant secretaries of State or Defense who had ignored Allen's staff quickly noted that Clark, a longtime Reagan confidant, enjoyed easy access to the Oval Office -- and, unlike most advisers, would happily take aides to presidential meetings. Suddenly, National Security Council staffers -- perhaps the least visible but potentially most influential bureaucratic players in the creation of U.S. foreign policy -- were no longer bypassed in the flow of memos and position papers that determines power in a bureaucracy.

Now the guard is changing again, as Vice Adm. John M. Poindexter becomes Reagan's fourth national security affairs adviser in five years and top aides are shuffled as well. In the shadowy world of the NSC, a new period of uncertainty has begun.

Congress created the NSC in 1947 as a high-level council -- the president, vice president, secretaries of State and Defense and the attorney general are the only members -- with a small staff to manage crises and knit a coherent policy for the government. Since then, it has soared to dizzying heights of power under President Richard M. Nixon and Henry A. Kissinger and tumbled to relative obscurity under Reagan and Allen.

Where Poindexter and his staff now take the institution could help determine whether the Air Force bombs Libya or exercises restraint, how the pain is apportioned in a new era of tight budgets and what will or won't be bargained away in negotiations with the Soviet Union.

Often the last advisers to whisper in the president's ear before he makes decisions, the little-known professionals of the NSC -- on loan from the Foreign Service or the military, from academia or the Central Intelligence Agency -- can wield considerable power from their niches in the Old Executive Office Building. They clear State Department cables and examine Defense Department speeches.

But the task of eliciting cohesion from a fractious administration is laced with frustration, a seemingly endless round of droning interagency meetings, drafting and redrafting option papers and cajoling or even pleading for information from better-known bureaucrats around town. The frustration has been particularly keen during the Reagan years.

"There's not many people above you, but there's nobody below you," said one current staffer who asked not to be identified. "You have to have contacts across the agencies, so you're never surprised when something crosses your desk . . . . And a good NSC action officer doesn't want to be perceived as being too powerful."

From the fateful moment in 1981 when the national security affairs adviser was moved from the White House's first floor to the basement, the NSC has had to struggle for a role. Buffeted by strong Cabinet secretaries at State, Defense and Treasury with easy access to Reagan, the NSC has often lacked the clout to resolve disputes or set consistent policy. Observers inside and outside government have rated the staff's quality somewhat below its predecessors'.

Those observers say the quality has improved since 1981 and that the NSC's influence gradually grew under Clark and during Robert C. McFarlane's recent tenure; the adviser and his deputy are back on the first floor. Backed by a deputy skilled in crisis management (Poindexter) and an executive secretary who acted as an efficient chief of staff, McFarlane was free to play a larger role with Congress and in arms-control issues.

Now, however, Poindexter has moved up to McFarlane's job, replaced by Donald R. Fortier, who is regarded by peers as intelligent and able but less experienced in crisis management. The executive secretary, Robert M. Kimmit, has moved to the Treasury Department, and his replacement is expected to leave soon.

Those losses, combined with Poindexter's relative inexperience in dealing with Congress and administration heavyweights, are likely to return the NSC to a period of diminished clout, at least temporarily, many insiders say.

"For the staff, the question will be, 'How will this affect relations with my opposite numbers in other agencies?' " said Geoffrey Kemp, an NSC staff director during Reagan's first term. "That is easily the most important barometer for the staff: Who will return my phone calls?"

Kemp and others agreed that, no matter who is adviser, the NSC must be reckoned with as long it writes a covering memo to almost everything the president sees.

"It is never true to say that the NSC is not powerful," a senior Defense official said. "It is powerful even at the worst of times."

But NSC staffers interviewed for this article said, often ruefully, that they rarely feel very powerful. One recently departed staffer recalled that morning meetings are dominated by discussion of what has been leaked to the morning newspapers, what will appear on the evening news and which Cabinet undersecretaries are battling.

"Ninety percent of the job is damage limitation and crisis management, and 10 percent creativity and planning," said Kemp, who served as the NSC's Mideast expert.

"Your day is very much dictated by foreign visitors, of which there is an endless stream," he added. "And if they're not coming into the White House, they're trying to get into the White House. I spent hours and hours at more lunches than I can count listening to the pleas of diplomats about why their officials should get an invitation.

"If you come from academe and think you're going to discuss the strategic problems of the Indian Ocean, you're wrong," Kemp said. "Instead, you listen to, 'How do we get our guy ahead of the guy from Kuwait?'

"And then," Kemp concluded, "you accumulate this enormous pile of reading material, which you learn to shuffle in almost inverse proportion to its substance. In academe, you fight for the Foreign Affairs [magazine] when the new volume comes in; here, it's always at the bottom of your stack, while you worry over the latest snippet of CIA gossip."

The NSC at times has played a more central role in making policy.

James V. Forrestal, the nation's first secretary of Defense, strongly supported the council's creation as a vehicle to control his boss, President Harry S Truman, whom Forrestal viewed as anti-military. Instead, the NSC evolved into a White House staff, used by presidents since John F. Kennedy to control the diplomatic and military establishments.

Nixon and Kissinger carried that trend to an extreme, keeping the bureaucrats busy in meaningless interagency groups while they made foreign policy from the White House, supported by a talented NSC staff. One senior Reagan administration official recalled that period with some wistfulness.

"There was none of this Cabinet government crap," the official said. "Maybe not all the decisions were wise, but at least there was decisiveness. Somebody was in charge of foreign policy. There was coherent, streamlined decision-making."

President Jimmy Carter maintained a strong NSC under Zbigniew Brzezinski, who strongly influenced foreign policy while protesting that he would not follow the Kissinger model.

"You were pretty sure you were at a very critical juncture in a very challenging process," said Roger Molander, whose NSC service spanned the Kissinger and Brzezinski years. "You were sitting inside the castle, you knew you were in a key position and people brought a lot of seriousness to the enterprise."

A favored tactic of that era was the option memo, prepared for the president by NSC staffers, with only one viable option. As Molander described it: "Do you want to kill your mother, do you want to kill your wife or do want to send the following message to the Russians."

"Sure, we could all plead a little guilty to doing that once in a while," Molander said. "But usually we would try to be honest."

The Reagan team came into office determined that the national security affairs adviser would funnel recommendations from the bureaucracy in an even-handed way rather than stack the deck.

"Any NSC has to play several roles," a senior NSC official said in a recent interview. "The one I would put at the top is being an honest broker for the system."

Unlike Brzezinski, the Reagan advisers have mostly stuck to their pledge to maintain a low profile and defer to the Cabinet. The New York Times index of 1978 could refer readers to 145 articles on Brzezinski, while in 1984 there were only 15 references to McFarlane.

Many of the foreign policy authorities who attacked Kissinger and Brzezinski for dominating the Cabinet call Reagan's advisers too weak.

"I find it very amusing when I hear nostalgia for the days of Kissinger and Brzezinski from the very same Georgetown set that was busily knocking Kissinger and Brzezinski a few years ago for usurping too much power," Brzezinski said in a recent interview. "I find it very amusing."

Brzezinski said in 1982 that the Reagan system was marked by "chaos and confusion." Since then, though, "it has settled down quite a bit," he said last month.

In some areas -- such as Central America policy, where the State Department has gone through three assistant secretaries -- the NSC has taken a lead. Administration officials said the NSC has also forged consensus on less visible issues that percolate in the bureaucracy's mid-levels, such as liberalization of trade with the People's Republic of China.

"We realized that Defense wasn't just going to have a technology-transfer policy imposed upon them," a senior NSC official recalled. "They had very legitimate concerns, so we had to sit down and talk through those concerns."

A steering group chaired by Fortier repaired to the White House situation room every three weeks -- and eventually more often -- to draft guidelines on what kinds of technology could be exported without endangering U.S. security. Assistant secretaries from Defense and Commerce squared off, joined by officials from State and Energy, the Joint Chiefs of Staff and NSC's Asian affairs director, Gaston J. Sigur Jr.

A working-level group of deputy assistant secretaries and office directors met twice a week during the same period, picking through thorny technical issues. After about six months, the bureaucracy had negotiated an agreement -- and was ready to negotiate with the Chinese from a mostly unified position.

"That's the trick," the NSC official said. "The hardest part of dispute brokering is building a strong consensus without watering down the policy to the lowest common denominator."

Without backing from a dominant adviser, the NSC action officer's job has called for more art than ever, several staffers said. Some become disdainful of the "bureaucratic pap" produced in the agencies, as one said, and lose their effectiveness, while others deal only with assistant secretaries, neglecting to cultivate the lower-ranking desk officers who may often be more knowledgeable.

"They think they're in the White House, and they can bludgeon the bureaucracy," one State Department official said. "It doesn't work that way."

Other staffers focus on who will accompany Reagan on his next trip or who gains entry to his office most often. But some glory in the nitty-gritty challenge of building allies across the bureaucracy.

Those who came to the NSC from one of the agencies are often at an advantage, some staffers say. One of the most respected assistants is Jack F. Matlock Jr., who heads the European and Soviet desk after holding similar jobs in the State Department, serving three times in the U.S. embassy in Moscow and representing Washington as ambassador to Czechoslovakia.

Phillip H. Ringdahl, who helped McFarlane fashion a compromise with Congress on economic sanctions for South Africa, joined the NSC as senior director for African affairs after almost 20 years in the State Department and in diplomatic posts in Africa.

Sigur, by comparison, joined the NSC from George Washington University; Poindexter is still an active-duty admiral; and Latin America director Constantine C. Menges arrived via the CIA.

"The best thing a staff officer can do is come up with a good idea and convince the department that it was their idea, because eventually it has to be implemented by the department," one NSC staffer said. "But is it easy for a Cabinet officer to block progress, to frustrate you? The answer is clearly yes."

That frustration has been clearest in arms control, where many battles -- reinterpreting the Antiballistic Missile treaty, adhering to SALT II, developing a small mobile missile -- have been fought to apparent conclusion and then immediately reopened. How do issues ultimately get resolved?

"They don't," said John Pike, an administration critic at the Federation of American Scientists. "NSC has simply fallen between the cracks, ground between State and Defense."

Arms control and other contentious issues that divide State and Defense, such as when to use military force, reflect a weakness in the NSC, one staffer said, but it is a weakness desired by Reagan, who wants strong players at the agencies. And to the extent that issues cannot be resolved, the staffer said, it is because of Reagan's "unwillingness to step in and make decisions."

"I assume [that] at some point, the president will make his choices, but I don't think he's had to yet," Brzezinski said. "Some basic decisions will have to be reached, and he will have to come down one way or the other. The stronger a National Security Council he has, the easier it will be.