Defense and space projects account for most increases in the $135 billion federal research and development budget next year, worrying scientists who fear that after years of growth the nation is beginning to skimp on technology that fuels marketplace innovation.

The realignment by Congress of research money toward national defense and human space exploration means many universities, institutions and scientists will have to scramble for new sources of money or cut back current or planned projects.

The National Institutes of Health, the nation's premier biomedical research agency, saw its budget doubled between 1999 and 2003 but is getting $28.6 billion next year, a slight 0.1 percent drop that marks its first budget cutback since 1970.

The cut, while small, comes at a time when a lot of research simply costs more, even the laboratory mice used in cancer research, explained Dr. Harold Varmus, a former NIH director and Nobel Prize winner.

"There is a battle for the future in science and technology. That's what is going to govern the future of our country. Not increasing investments in those areas sends a signal the country is going to regret," said Varmus, who now heads the Memorial Sloan-Kettering Cancer Center in New York.

The Bush administration counters that federal research and development spending remains near an all-time high and is close to 45 percent higher than when the president took office.

That "is a strong statement of the high priority this administration places on innovation, competitiveness, science, technology and research," said Donald Tighe, a spokesman for the Office of Science and Technology Policy, which advises the president.

Federal research and development spending will rise $2.2 billion, or 1.7 percent, in 2006, to about $135 billion, according to an analysis by the American Association for the Advancement of Science.

Of that increase, 97 percent will go to Department of Defense weapons development and National Aeronautics and Space Administration spacecraft programs, AAAS said.

Funding for other federal R&D increases only slightly, and actually falls if adjusted for inflation, AAAS analyst Kei Koizumi said.

"For 2006, for most areas, it's looking pretty bad. The total is going to be a new record, but it's going to be big increases in two areas," Koizumi said. "Obviously, those are big priorities but in an overall budget in which Congress and the president are trying to cut domestic spending, all other R&D programs are flat at best and falling in most cases."

The nation's universities and research institutes fret the emphasis increasingly falls on development, which tends to help industry, instead of the experimentation and exploration associated with basic research.

Research spending is falling or stagnating, disproportionately hurting the colleges and universities that depend on federal support to run their electrical engineering, computer science and other departments, said Tobin Smith, senior federal relations officer for the Association of American Universities.

The group's 60 research universities account for 60% of federally supported, university-based research.

Most immediately, decreased R&D spending will lead to layoffs and other cutbacks at some facilities that rely on federal funding, including the Department of Energy-supported Relativistic Heavy Ion Collider at Brookhaven National Laboratory.

The Long Island particle collider creates subatomic head-on collisions between intersecting beams of gold ions, allowing nuclear physicists to study what conditions prevailed at the birth of the universe.

A roughly $40 million cut to the center, coupled with higher electricity costs, will force the collider to shut down for the year, said Michael Lubell, a City College of New York physicist and spokesman for the American Physical Society. About 100 workers will lose their jobs, he said.

"Once you go down that road it is very difficult to make a U-turn and go back to where you were," Lubell said.

NASA will see a 7.3 percent increase in R&D funding, much of it for spacecraft to carry humans to the moon and beyond. R&D spending on the departments of Homeland Security, Transportation and Interior, including the U.S. Geological Survey, will rise. But spending on Energy, Commerce and Agriculture will drop, as it will for the NIH, AAAS said.

Promoters of basic research said investments made today may not pay off for years, if not decades. The National Science Foundation notes it supported Google Inc. co-founders Sergey Brin and Larry Page back when they were graduate students at Stanford University studying better ways of searching for information on the Internet.

"A lot of innovation comes from basic research, and it takes a number of years for that basic work to transform itself into innovations," said Sam Rankin, associate executive director of the American Mathematical Society and chairman of the Coalition for National Science Funding.

Others say the increased Pentagon spending should help the economy, producing more civilian spinoffs than ever before as spending shifts from large weapons systems toward networking and information-processing technologies.

"I would say Pentagon spending on research and development is more robust today than at any other time since the 1980s," said Loren Thompson, a defense analyst with the Lexington Institute think tank in Arlington, Va. "In addition, it is scattered among more projects that have potential civilian applications."