BEDFORD, MASS. -- A sprawling maze of buildings sits perched at the top of a broad driveway that winds through meticulously manicured grounds. Gathered inside are thousands of engineers, computer scientists, mathematicians and other specialists who are supposed to serve as the Air Force's high-technology seers on air defense and communications projects. At another site in McLean, more than a thousand other scientists toil away on a range of projects from communications and intelligence systems to air traffic control and computerized office equipment. The modernistic building in McLean that houses the Washington operation and the Bedford complex -- with its marble, ficus trees and sunlit atrium -- have the same dressed-for-success look found at many defense contracting companies. And with 1988 revenue of $463 million, and $62 million tucked away in a capital reserve, this is an operation that would be the envy of many a Beltway company scrambling for customers and contracts. But this is not just any high-tech enterprise. It doesn't worry much about attracting customers or maximizing profits. Its primary customer, the federal government, never fails to pay its bills. And without so much as lifting a marketing finger, it has the various agencies of government falling over themselves to use the company's services. Welcome to Mitre Corp., where the best of government and industry are supposed to come together to lead the federal government through the thicket of technology choices. To its supporters, Mitre provides an elegant solution to a number of problems vexing the government, among them the federal brain drain, the many conflict-of-interest controversies that surround for-profit companies, and the difficulty the government sometimes faces in getting consistent, high-quality work. "I don't want to knock for-profit companies because they are the backbone of the success of the United States," said Charles S. Zraket, president and chief executive officer of Mitre. But at a for-profit company, "the executive's prime concern has to be: Are they going to make a profit out of this or lose a lot of money... . I happen to be a person who wants to optimize around the quality of the job I am doing. It would be very difficult to build an organization of the quality and independence that Mitre has and have it be a profit-making organization. The two are incompatible," Zraket said. Mitre's critics, drawn mainly from the government services industry, see a darker side to Mitre. They contend that Mitre is grabbing work that rightfully should be done by the private sector. And they charge that Mitre, without the financial discipline of a for-profit corporation and the market discipline of having to compete for business, has grown fat and inefficient. Industry executives also complain about Mitre's lack of public accountability. There is no congressional oversight of Mitre's activities, and its salaries -- which are reported to range from competitive to generous -- are neither regulated by civil service law nor available to the public. And industry executives also suspect that the reason federal agencies seem so enamoured of Mitre is that it gives them an easy way around the government's tedious and time-consuming competitive bidding process. "Mitre has played the game to the hilt in using their preferred status to do things to get work, and to grow in terrible competition with the private sector," said Robert W. Krueger, president of Professional Services International, a Los Angeles consulting firm. There is no disagreement that Mitre has grown into a powerful voice in the government since its founding in 1958, influencing the direction of government research and purchases in many areas. It decides what technology the government should pursue in a variety of areas from communication to computer systems. In many cases, it develops the preliminary requirements for projects, designs the prototype for the projects, helps review proposals from for-profit contractors, and -- once the contract is finally let -- monitors how well the work is carried out. This power to decide the fate of thousands of companies has created a strange environment, in which executives of private firms bash Mitre in private while publicly courting its approval. The Professional Services Council, for example, representing government contracting firms, has complained bitterly about the growth of Mitre and formed a committee to fight such non-profit competition. But when called for this article, the chairman of the committee refused to say a word about Mitre once he realized his company's name would be associated with his comments. Other executives agreed to talk, but only on the condition that their names not be used. While Mitre is among the oldest and most celebrated of the government's think tanks, it is not the only one. There are 36 non-profit groups -- known as Federally Funded Research and Development Centers, or FFRDCs -- that receive close to $6 billion in federal contracts, or about 10 percent of the government's research and development budget. Many of the FFRDCs were formed in the years after World War II, when the military, embarked on a technology race with the Soviet Union, found that there were few private companies offering the expertise they were seeking. But in recent months, there has been a flurry of new interest in using the FFRDC's as agencies struggle to sort through the vast array of new technologies being offered by private industry. In May, for instance, the Army announced it would establish a new FFRDC on Electromechanics and Hypervelocity Mechanics. Then a few weeks later, the Office of Management and Budget said it was considering creating an FFRDC to provide federal agencies with technical advice on the management of major government information technology systems. The Social Security Administration recently turned to the Institute for Defense Analyses to help it plan for its large computer capacity needs in the 1990s. And earlier this summer, the head of Pentagon's operational testing office decided to shift work from two defense contracting stalwarts -- TRW Inc. and Science Applications International Corp. -- and transfer it to IDA. "We anticipate that over the horizon that there will be major changes in computer systems," said Jaime Manzano, director of the office of strategic planning at the Social Security Administration, which has struggled for several years to improve its outdated computer systems. He said the agency wants to be able "to anticipate future technological change rather than being overtaken by events." In the case of Social Security, the agency's officials concluded they needed a group that could approach several computer companies, examine their proprietary products, choose the combination of technologies that best fit the agency's needs. A for-profit company would not have had access to proprietary data and the inside thinking of company executives, said Manzano. Other government officials said they turn to FFRDCs because they are more responsive to their needs. Veda Inc., an Alexandria-based firm, got the idea of using concentrated solar energy to destroy hazardous waste. It hooked up with Babcock & Wilcox, an energy engineering firm, and got a Department of Energy grant to do a $50,000 feasibility study. Using Veda's concept, Babcock & Wilcox put together a design for a solar reactor in which hazardous PCBs could be destroyed. As a consultant, Veda brought in the University of Dayton Research Institute, which had worked extensively with PCBs to study any potential problems with the reactor. It was all going nicely for Veda until, after the initial study, the Energy Department transferred the work to the University of Dayton Research Institute and the Solar Energy Research Institute, an FFRDC based in Colorado. E. Neville Hunter, Veda vice president, said the results coming in are confirming Veda's predictions, yet for now the company remains out of the project, much to its frustration. Frank Wilkins, a solar energy program manager at DOE, said he decided to switch to the university research institute and SERI because he was concerned by the for-profit companies' decision to focus on the design of the solar reactor rather than concentrating on basic research. He said he tried to change their focus, but to no avail. "Generally, industry does everything toward the end of having a product they can sell relatively quickly," Wilkins said. In another case, the Army turned over work to Los Alamos National Laboratory that had been proposed by and awarded to Geospectra Corp., a small image-processing firm in Michigan. Geospectra had negotiated with the Army for more than a year, but because of the lengthy procurement process, the Army was afraid it would not meet the fiscal year deadline and would lose the funds for that year. So they canceled the competition and sent the money and the work to Los Alamos Lab. Much of these same attitudes were discovered by the General Accounting Office, which investigated the use of FFRDCs. A GAO report issued last year found that Defense Department officials favor FFRDCs because they "believe that FFRDCs are especially competent and convenient, while industry is believed to sometimes lack objectivity and, because of lengthy government procurement procedures, to be relatively inconvenient." These attitudes among government officials have made the fight to reduce the work done by FFRDCs an uphill battle for industry, all the more so because many of the federally funded research labs and think tanks are prestigious and well-respected institutions with strong and loyal followings. They include the Lincoln Laboratory in Massachusetts, Lawrence Livermore National Laboratory in California, Plasma Physics Laboratory at Princeton, N.J., the Oak Ridge National Laboratory in Tennessee, Aerospace Corp. in California, Rand Corp. in California, and three Washington-area based outfits -- the IDA; the Logistics Management Institute, and Center for Naval Analyses. Mitre, however, has emerged as one of the most controversial. Industry executive acknowledge that, ironically, this animosity is partly a reflection of Mitre's success. "They act more, in the marketing sense, like a private sector organization than any of the other FFRDCs. That is certainly a compliment to ... their management ability," said Krueger, who founded Planning Research Corp. and headed it for 19 years. "I think Mitre has good people and they do good work. I understand why the government uses them," agreed Wayne Shelton, the current president of PRC and himself an alumnus of the non-profit Rand Corp. But Shelton added that Mitre should be far more limited in the scope of work it can do. Mitre is headed by Zraket, known as Caz by friends and colleagues -- an electrical engineer who has been with the organization since its formation. Zraket is described as a man who loves technology and who is both credited and blamed with Mitre's healthy growth over the years. Yet even those who have spent years battling Zraket and Mitre have few harsh words for him personally, and praise him for his accessibility and knowledge of the subject. "We disagree on a lot of things, but in general Caz is one of the most forthcoming and straightforward people I know," said Gary Chapman, executive director of Computer Professionals for Social Responsibility, which seeks better control over the technology used in the nation's nuclear arsenal. "Obviously, he has more faith in the benefits of technology in the military than I do. But I think he is extremely smart, a real scholar in many fields, in addition to being an excellent manager." When Mitre was formed in 1958, as a spinoff from the Lincoln Laboratory, the Air Force desperately needed a company to design its air defense system and could not find a for-profit company to do the job. "They wanted a dedicated laboratory that had the multi-disciplines of radar, computers and communications. It was a new technology, and there was no expertise for this anywhere," said Zraket. Soon Mitre played a key role in the development of numerous air defense systems for the United States and other allied countries, helping design the electronic brains of systems that detect and track enemy missiles and aircraft, and intercept enemy communications. Mitre played a major part in designing the hardened, underground facility at Colorado Springs to house the North American Air Defense (NORAD) system for both air and ballistic missile attack. Mitre is currently assisting the Air Force in modernizing the Airborne Warning and Control System, or AWACS. Mitre also got heavily involved with satellite communications technology, and is currently working with the Milstar satellite communication system. Milstar, a system designed to survive nuclear attack and enemy efforts to jam it, is intended to provide worldwide communications for the military. One of Mitre's tasks is to use simulation to study the effects of nuclear disturbances on Milstar signals. During the Vietnam War, Mitre became a target for antiwar activists for its role in developing an "electronic fence," composed mainly of acoustics and sensors, that was supposed to help pinpoint the movement of the Viet Cong and North Vietnamese troops. To the war's critics, it was a symbol of all that was wrong with the American strategy, an expensive, high-tech answer to a war that required low-tech solutions. Mitre also does plenty of non-military work. In its early years, Mitre became involved in air traffic control work for the Federal Aviation Administration, an extension of the air defense work it was doing for the Air Force. It is currently advising the FAA on the massive upgrade of the nation's air traffic control system that is to be completed in the 1990s. Also receiving Mitre's advice these days are the General Services Administration with FTS2000, the project to replace the federal telecommunication system; the Securities and Exchange Commission, which is designing a new computer system, the National Institutes of Health, which is upgrading its Medical Literature Analysis and Retrieval Systems; and Georgetown Hospital, which hopes to create a filmless hospital that would store radiological images on computer tape rather than on film. Mitre is pushing into new areas as well, exploring sensor technology to cope with radar-evading stealth aircrafts and the emerging neural network technology that models computers after neurons in the brain. The military projects are handled under the so-called "C3I division," which has about 3,000 professional employees. The non-military projects are handled by the civilian division, a non-profit operation with about 800 professional employees who are based in McLean. Zraket has heard the criticism about Mitre grabbing business that should go to industry. But he said industry's anger is misdirected. "The government is really the one who decides what we do," he argues. "We don't bid on anything." Executives, however, cite example after example of times they made a proposal for a study, only to be told that Mitre had decided that it should perform the study instead. The executives, however, declined to discuss those incidents for the record. One instance, cited frequently, was Mitre's help in designing the FBI's National Crime Information System. Zraket contends that Mitre tried to reject the work, knowing that it would generate industry criticism, as did the Air Force, which must pass on work done for other agencies by the C3I division. But those decisions were eventually overturned after the FBI appealed to an assistant secretary of defense. Losing contracts to Mitre particularly infuriates people plying their trade along the Capital Beltway, because they contend that higher salaries at Mitre make it a more expensive choice for the government. Zraket and other Mitre officials strongly dispute that accusation, saying most salaries are equivalent to or below industry averages. Blake Smith, an official with the Air Force's Electronic Systems Division, said that because of complaints his agency regularly reviews Mitre salaries but generally finds very few out of line. "Mitre is in a competitive market, so we have to recognize the market they are in when they are trying to draw and retain people," Smith said. "We want them to have the talent necessary to do the job. To do that, they have to have access to the best minds." Smith said the Air Force does not object if salaries are slightly above the general industry level, but becomes concerned if they are significantly higher. Zraket, for his part, said the criticism of salaries "frankly infuriates me. It comes from companies where executives are making two and three times the salaries of Mitre executives." Zraket, however, declined to reveal his own salary. And unlike a publicly traded corporation, or a government agency, Mitre is under no obligation to do so.