THE U.S. SPACE AGENCY has built its reputation by doing the impossible. But it has been spread so thin for years that sometimes it has trouble doing the merely necessary. That's what happened this summer.

In both the Hubble Space Telescope and the space shuttle programs, the fancy stuff -- the complex software and high-performance machinery that made each unprecedented -- is working pretty well. The problems that caused a furor over the past few months were of the mundane sort. Still, they made the overall program seem a shambles.

The rash of bad news has obscured the fact that the system seems to be slouching toward long-needed changes. Among other things, a blue-ribbon panel next month is expected to deliver recommendations to NASA and the White House that could help the space program clear unwanted baggage from the past and bring the future into focus.

Norman Augustine, chairman of the Advisory Committee on the Future of the U.S. Space Program and chairman and chief executive officer of the Martin Marietta Corp., opened the panel's hearings in September by pointing out one of its biggest dilemmas: While there is broad support for a space program in general, "it seems that there are no two people who support the same space program."

The public at large, in the view of some, still does not understand the difficulty, cost and risk of space activity.

"There are too many people who want a voodoo space program -- you know, 'Beam me up, Scotty,' " said J. R. Thompson, second in command at NASA and a survivor of many struggles with difficult projects. "It just takes a lot of hard work . . . . I am as disappointed as anybody over the Hubble. However, I am in awe of what our scientists and engineers have accomplished."

Like the 1986 Challenger disaster, caused by the failure of a simple joint, the recent embarrassments have reminded everybody that the space agency can accomplish miracles of science and technology and still be brought low by a human error measured in thicknesses of a hair.

In the Hubble case, for example, one day 10 years ago a technician apparently used a simple measuring wand improperly -- possibly upside down -- leading to a heartbreaking flaw in the telescope's primary mirror which managers then failed to detect. The hydrogen leaks that halted shuttle flights for several months were caused by faulty workmanship, inadequate tests and other consequences flowing from such simple errors as allowing sandpaper grit and microscopic glass beads into fuel lines.

Now the shuttles have begun to fly again, and NASA has secured a 13.5-percent increase in its budget for 1991 despite the snarl over the federal deficit. But the real problems run deeper. Some are inside NASA. Some are caused by the larger political system in which it operates, which to some extent ties the agency's hands.

"NASA is the most difficult management job in the world, bar none," Erich Bloch, former director of the National Science Foundation, told the Augustine panel. The agency today has to manage many more projects, with far more complexity, than it had during the glory days of Apollo -- "too many," according to Bloch. But it has a much smaller budget and fewer people.

NASA's inability to prevent or catch simple but devastating errors has re-ignited doubts at both ends of Pennsylvania Avenue about how fit it is to storm the barricades of the future.

The stakes include a $37-billion manned space station, a $30-billion Earth-observing satellite system, ambitious robot missions to study the heavens, and exploration of the moon and Mars by humans and robots at a cost that could total hundreds of billions over several decades. Setting Fair Standards Complaints about the agency that surfaced after the Challenger accident have been revived and expanded: NASA has become a fossilized, overcautious bureaucracy with an over-the-hill workforce that has lost its capacity for bold innovation; it is not equipped to manage large complex programs or to argue credibly for the space agenda on Capitol Hill; and it is addicted to mammoth showcase projects at the expense of its primary mission as an innovator in science and technology. And so on.

Some critics have suggested that NASA be broken up and restructured; some urge that the space program be divided up among several agencies more than it already is.

But others say they are impressed that the agency has accomplished as much, and attracted and held as many talented scientists and engineers as it has, given the constant struggle to get projects funded, the long years they require to reach fruition and the low pay, bad press, red tape and other problems of government work.

The Orlando Sentinel, whose circulation area includes NASA's Kennedy Space Center, wondered recently why NASA suffers the torments of political hell for its delays and blunders, while other agencies do not. "The U.S. Agriculture Department reported that at least 13 of its employees died on the job last year, and no one blinked," the paper pointed out. Seven astronauts died in the Challenger accident, which grounded the shuttle fleet and inspired warnings that one more such tragedy will end the manned space program.

The Air Force, which last year witnessed 76 airmen killed in plane accidents worldwide, has been rocked by weapons-procurement scandals involving dollar amounts that dwarf the civilian space program. And in the military space program this year, it dropped a Titan rocket motor, killing a technician. In addition, the Titan 4 rocket, the unmanned replacement for the shuttle for most military payloads, has suffered delays and technical setbacks. But its viability as an agency is not in doubt.

What makes NASA different? One factor, says policy analyst John Logsdon of George Washington University, is that the public and Congress support the space program "not only for its intrinsic value but as a symbol of what we can do well." Stuck With the Shuttle The fundamental weaknesses of America's civilian space program are no mystery to anyone who cares. Its underpinnings are absurdly fragile. It lacks a reliable ride to work -- that is, a versatile fleet of economical manned and unmanned space vehicles that can fulfill the basic requirement of ready access to space.

Forced dependence on the shuttle -- a policy discredited by the Challenger accident -- has skewed everything from the design of the Hubble telescope in the 1970s and the space station in the 1980s to the preoccupations of NASA's leaders in the 1990s.

For instance, the telescope's planned size was reduced so that it would fit into the shuttle cargo bay and its orbit was lowered from 22,300 miles, where it would have had an unobstructed view, to 380 miles -- just under the shuttle's maximum delivery altitude -- where the Earth will always block part of its view.

The shuttle cargo bay and other shuttle specifications have been tailored to accommodate military performance and size requirements for transporting heavy spy satellites to orbit. Ironically, the Pentagon has now decided to transfer the job of launching spy satellites to unmanned rockets. Since Apollo, the program has lacked a well-defined, compelling goal -- or even a ranking of its priorities. Eager to please all its constituencies, the agency has kept projects going for each of them but has been forced to allow them to lag behind schedule, sometimes consuming huge chunks of scientists' careers with the wait. The description heard often on Capitol Hill is that the space program is "10 pounds of potatoes {or some other substance} in a five-pound sack."

(The only NASA project in recent times to run ahead of schedule and below budget is the new orbiter to replace Challenger. What made the difference, officials say, is that, since a proven design was involved, NASA was given a lump sum of money in advance, rather than having to fight constant skirmishes for it. The agency has appealed for more stable, predictable funding for other projects.)

There is considerably less agreement about how to repair these shortcomings. Somebody has to cut the Gordian knot and decide. As the budget mess has shown, even a crisis is not necessarily enough to wring decisions out of the current system. But there are signs of improvement. The president took a step toward defining the program's goals for the future with his moon-Mars exploration initiative. But the proposal so far remains remote, vague and unfunded by Congress.

That leaves the proposed international space station Freedom, in all its confusion, as the most tangible embodiment of NASA's vision of the future over the next 40 years. {See box below.} But it also has in its portfolio the vastly complex Earth Observing System. As part of an international plan to study how Earth is changing -- in order, presumably, to save it -- the project is politically popular but involves immense costs, and data-collection and management challenges that some say have been vastly underestimated.

The Augustine panel, meanwhile, has been getting an earful of advice from all sides. Spurred by the Bush administration's commitment to the manned space program, the committee is viewed by many in the aerospace community as "the last, best hope" the program has to break out of the funk it has been in for two decades.

"We're in the process of lowering expectations to match rising -- though not rapidly rising -- resources," said Logsdon. "The right thing to observe about the NASA budget this year is its overall increase." The Cost of Competence Prime targets for the panel are likely to be the management and people problems that underlie all NASA's undertakings. Its sprawling system of specialized facilities that sometimes duplicate work or compete with each other and its shortage of people to monitor contractors adequately are examples. And over the last 20 years, NASA's burden of federal red tape -- procurement regulation, civil-service rules and the like -- has grown along with everybody else's. Its ability to compete with the private sector for talented scientists and engineers has eroded.

"When dealing with tens of billions of dollars over tens of years, the idea that these {programs} can be managed on the public's behalf by less than the best people is an absurdity," said John McElroy, dean of engineering at the University of Texas.

But "the government is not even remotely competitive," he said. "I can't hire a chairman of a department to manage a million-dollar budget and eight people for the salary we currently give the administrator of NASA." And that inability to compete for top talent, he said, "cascades down to the engineers and managers below."

Whatever the changes, it seems likely that in order to win support, NASA still must do business with essentially the same system of pork-barrel politics, conflicting science and technology interests, bureaucratic turf wars in the executive branch and geopolitical strategizing that shaped the Hubble Space Telescope, the space shuttle and the space station.

Harrison Schmitt, former astronaut and U.S. senator, lamented that "few in Congress currently benefit politically from support of manned space flight," because it has no organized grass-roots constituency.

NASA must also deal -- as both partner and competitor -- with increasingly aggressive space programs in Europe, Japan and other nations.

Under the 1958 law that created the space agency, and the laws of physics that govern its science and engineering, NASA's performance is unusually visible and quantifiable for government work.

"These guys have their pants down around their ankles every day," an administration official who has been critical of NASA said in a sympathetic moment. "They don't have the skirts and veils that Defense has, and those really help sometimes."

These are tough times for romantics, and a lot of the space program's appeal is romantic. But author Daniel Boorstin, former director of the Library of Congress, went against the spirit of the times in his testimony before the Augustine committee. He called space "the great treasure house of the unknown" and warned against a public "obsession with the predictable" and against "the cost-effective syndrome," which seeks to avoid risk-taking by requiring enterprises to justify themselves with advance assurance of a certain level of payoff.

Such measurements, Boorstin said, "are difficult to apply to the great enterprises of exploration and discovery . . . . The most wonderful things in life are not cost-effective -- like love and children."

Kathy Sawyer covers space for The Washington Post.