A government regulation that would further reduce emissions from diesel cars by 1985 is not warranted now and could prove to be a "costly" mistake, according to a yet-to-be-released study by the National Academy of Sciences.

A 20-member academy committee did conclude that the existing standard for the soot, or particulate matter spewed out by diesel cars -- now in effect for 1982 models -- is "practical and prudent" for protecting human health and the environment.

But the group says that more time is needed to resolve "uncertainties" associated with the more stringent standard set to go into effect for the 1985 model year. The uncertainties include possible health effects, the feasibility and cost of achieving the standard and the market for such cars.

The report is expected to play a major role in the Environmental Protection Agency's consideration of an industry petition to suspend the 1985 standard, as well as congressional efforts to revise the Clean Air Act. The report is likely to be "controversial," admitted one committee member.

The committee also urged the EPA and Congress to consider regulating heavy diesel trucks and buses, which release larger amounts of dirty particles into the air.

"This may be more cost-effective than tightening the emission levels of diesel cars and light trucks," the Diesel Impacts Study Committee contended.

The panel, chaired by Stanford University professor Henry S. Rowen, outlined several alternatives for regulating diesel cars.

These include setting the 1985 standard at an "intermediate" level, giving manufacturers greater flexibility through "averaging" emissions levels of their fleets, instituting state standards to allow stricter requirements for areas with particular pollution problems and levying fees tied to the level of the emissions.

The findings are contained in a summary of the final report recently circulated to members of the committee. A copy was obtained by The Washington Post and the findings were confirmed in interviews with several sources familiar with the committee's deliberations.

The long-awaited study will be delivered soon to the federal agencies that ordered it, including the White House Office of Science and Technology Policy, EPA, and the departments of Transportation and Energy.

Natural Resources Defense Council attorney David Doniger disagreed strongly with the findings, when they were described to him. "We firmly believe that the 1985 standard is achievable and that it is needed to reduce the risk of lung cancer and other respiratory diseases and the very likely consequence of a gross decline in visibility in urban areas," he said.

General Motors spokesman William Chapman expressed satisfaction that the committee members had "done their work exceptionally well" and restated the auto industry position that it was not possible to meet the 1985 standard.

Stanley Blacker, an EPA official, said the academy report could play a "significant" role in agency deliberations because it provides a "useful data base."

The study, begun more than two years ago, was a response to concerns about the potentially rapid growth in diesel cars. More than 4 percent of new cars are diesels; estimates suggest less than 2 percent of the total cars on the roads are diesels.

Although diesels are more fuel-efficient, they emit from 30 to 100 times more soot than gasoline cars, with the microscopic particles carrying a variety of toxic chemicals.

EPA established standards in 1979 to control these emissions, with an initial reduction to .6 grams per mile and then a later one to .2 grams per mile. The agency later postponed the standards to 1982 and then 1985.

A proposal to regulate emissions from heavy-duty trucks and buses, prepared during the Carter administration, is still pending.

The academy study sought to examine the technological, environmental, health and policy aspects of regulating diesels. The findings, outlined in the available summary and in interviews, include the following:

Control technology. The 1985 standard would require the addition of an exhaust after-treatment device, such as a "trap oxidizer," on larger cars. The committee concluded that it will not be technologically possible to get a proven device ready in time to meet the 1985 standard. One EPA official noted, however, that the National Academy of Sciences data was gathered in 1980 and there had been "substantial" advancements in the technology since then.

Health and environmental effects. The committee found no convincing evidence, said the summary, that exposure to diesel fumes causes human cancer or lung disease, but more research was urged. Major, long-term growth in the diesel fleet could have a large impact on the visibility in urban areas, it says.

Cost-benefit estimates. Although compliance for cars through 1984 is technologically feasible at less than $30 per vehicle, the committee said that costs of achieving the 1985 standard are "considerable"--from $150 to $600 per car.

The added cost might deter some buyers, the academy analysts suggested, and that would mean the loss of diesel benefits such as lower fuel consumption and passenger safety (because the cars are bigger than gasoline models with the same fuel efficiency). But the committee found that "uncertainties" in documenting health risks and other factors made it difficult to do a standard cost-benefit analysis.

Instead, a committee panel chose a "worst case" scenario for each option. The academy experts calculated that the costs of meeting the 1985 standard might turn out to be $2 billion to $3 billion annually, compared with $80 million to $1.2 billion for the 1982 standard.

The committee therefore advised waiting two to five years so that a "more informed" decision can be made, with the need for more stringent standards reevaluated in 1983 and every three years thereafter.

The committee was composed of 20 members drawn from diverse fields, with advice from four panels. Although emphasizing that he would "not charge the committee of being biased in terms of makeup," environmental attorney Doniger did charge that the "process was biased" because there was not an adequate opportunity for public input.