Development of nuclear energy has been stymied for decades on Earth because of high costs, the fear of another accident like those at Three Mile Island or Chernobyl, and the problem of how to deal with radioactive waste. But NASA and planetary scientists see it as opening a new era of research in space by providing a plentiful power source for deep space probes, which has been lacking so far.

"This is an unprecedented opportunity for exploration," said Dr. Ronald Greeley, a professor of geological sciences at the Arizona State University. Dr. Greeley is co-leader of a team of 38 scientists who have been working with NASA since February to define what kind of science could be pursued by a nuclear-powered spacecraft at Jupiter.

The proposed spacecraft, the Jupiter Icy Moons Orbiter, would be able to carry new and more powerful instruments and would be able to move in and out of orbit of different Jovian moons.

The team of scientists presented its recommendations at a meeting of the American Geophysical Union here last week. Still in the preliminary planning stages, Jimo would cost several billion dollars, which includes developing a space-worthy nuclear reactor. The earliest launch date would be 2011, and, Dr. Greeley said, "None of us will be surprised if it launches later than 2011."

Jimo would also have to overcome concerns about what could happen in case of an accident.

As many as 1 in 10 rocket launches still fail, opening the possibility of a nuclear reactor exploding high in the atmosphere and dispersing radioactive material. "As you introduce more nuclear power into space missions, you're looking for trouble," said Bruce K. Gagnon, coordinator of the Global Network Against Weapons and Nuclear Power in Space.

A space probe like Jimo would also be "an icebreaker to institutionalize nuclear power in space," Mr. Gagnon said. "Later it would be used for military purposes" like powering space-based lasers, he said.

NASA officials said the nuclear reactor would not be turned on until after the spacecraft reaches orbit and that safety would be a primary design concern. The nuclear fuel will be designed not to break up even if the rocket explodes.

The appeal of a small nuclear power generator for propulsion in space is that power becomes more precious with distance from the Sun. In the outer solar system, sunlight is too dim for solar panels and technologies like fuel cells are not reliable enough for the years-long voyage to distant plants. Until now, NASA has used hunks of radioactive plutonium that generate heat as they decay, and the heat is converted into modest amounts of electricity.

NASA's Galileo spacecraft, which in September concluded a hugely successful 14-year mission with a deliberate plunge into Jupiter, subsisted on an electrical diet of a few hundred watts - which would light up a handful of light bulbs.

That was still enough for Galileo to produce a wealth of discoveries, especially about Europa, one of Jupiter's moons. Cameras captured up-close images of Europa's exquisitely cracked crust of ice, and magnetic field measurements indicated that beneath that ice is a layer of electrically conductive material - what planetary scientists
believe is a liquid ocean and possibly the most likely place in the solar system to find life. Because Galileo had only small maneuvering thrusters, it could not enter orbit around any of the moons. Galileo's brief flights past Europa produced high-resolution images of less than 0.1 percent of the surface.

With nuclear propulsion, mission controllers would be able to drop into orbit around one moon for several months, then restart the engine to propel the probe to the next destination. "You are in the driver's seat," said Raynor L. Taylor, program executive for icy moons orbiter project. "You can control where you can go."

With a reactor on board, the spacecraft would also have perhaps a thousand times more power available. That would make possible power-hungry instruments like ground-penetrating radar to look at what lies beneath the surface. Scientists would like to investigate the geological history of the moons' surfaces and identify the types of molecules there, especially ones that might indicate life.

Nuclear propulsion is an old idea. In the 1950's, the United States attempted to develop a rocket propelled by small atomic bombs tossed out the back. NASA now envisions something less dramatic.

A uranium fission reactor, which would not be turned on until the probe reaches space, would generate heat that would be converted into electricity to power an ion engine. Ion engines use electric fields to accelerate ions to very high speeds, producing thrust.

NASA successfully tested a nonnuclear ion engine on an experimental spacecraft called Deep Space 1 in 1998. The ion engines in Jimo would have to be at least 10 times more powerful. NASA's Glenn Research Center in Cleveland announced last month a successful test of one possible design.

Sean O'Keefe, NASA's administrator, originally pushed for sending a nuclear-powered spacecraft to Pluto, the only planet not yet visited close-up. However, planetary scientists rallied to the defense of a lower-cost, conventional space probe already in planning. There is a rush to reach Pluto, which is cooling as it moves away from the Sun, before its atmosphere freezes and disappears.

The scientists working on Jimo are much more enthusiastic about the promises of nuclear propulsion. "This would be taking steps in leaps and bounds potentially," Dr. Greeley said.

Conceptual designs envision a spacecraft about 100 feet long with the nuclear reactor at one end and the scientific instruments at the other. Three studies for more detailed designs are under way.

"We have been surveying industry for their technology and capabilities," said Mr. Taylor of NASA. "We see no roadblocks."