An agency of the U.S. Department of Transportation is seeking approval to take random blood samples and urine specimens from "volunteer" motorists to determine whether drugs play a significant role in traffic safety.

The problem with the proposal at least as it was seen in the office of Transportation Secretary Brock Adams, was that there was some question as to just how much free will the volunteers would really have.

"In the document I saw," spokesman Robert Holland said, "it mentioned that the policeman would pull the driver over . . . Even if it's a good purpose, you don't intimidate people like that."

Dr. Monroe Snyder, who works in the office of driver and pedestrian research at the National Highway Traffic Safety Administration, where the proposal originated, said, "We can't do a study where people don't have privacy, and we're not going to."

Snyder said that he did not know where the impression of specific police involvement found its way into his proposal, which he said was forwarded in mid-September in the first step of a 12-step approval procedure.

"Since that package was sent upstairs," he said, "we have been asked to provide additional information to our administrator." Joan Claybrook. "That is all I know about its present status."

The proposal is in the form of a request to seek authorization for a consultants' study on drug problems in driving. According to Snyder it says the consultant "shall specify the sampling procedure to be used in collecting the data" including such matters as where, how and by whom a motorist is stopped.

Before any consultant would be permitted to go forward, his methodology would be carefully reviewed, Snyder said.

"The objective we have is to determine whether drugs are a significant safety factor," Snyder said. "If they are, should the federal government get involved? If not, fine. But it is not a simple question to answer. When you say 'drugs,' you're talking about thousands of possible compounds. It's mind-boggling."

Alcohol is a known factor in a very high percentage of auto accidents, especially fatal accidents. But it was years before research was completed that definitely nailed that fact down. Now most states require that drivers agree that they would willingly submit to a test for alcoholic content in their blood under certain circumstances as a prerequisite to obtaining a driver's license.

There is no such provision for drugs, and ther is little date. It was to fill this void, according to Snyder, that a program is being proposed.