The Environmental Protection Agency has launched an intensified investigation to see if 10 miscarriages among eight women in a tiny Oregon mountain town were caused by herbicide spraying in nearby forests.

In announcing the probe yesterday, EPA Administrator Douglas Costle said the Agency has already conducted preliminary investigations, but "we're uncertain at this point whether herbicides are responsible for the miscarriages."

"But," Costle added, "one of the herbicides in question, namely 2,4,5-T has caused birth defects and still-births among laboratory animals such as mice, rats, hamsters and birds. Another of the brush killers, Silvex, is being studied by EPA for the same reasons."

The problem was brought to the government's attention in April when schoolteacher Bonnie Hill and seven other women from Alsea, Ore., wrote EPA to say that they had found "an incredibly close correlation" between the dates of their miscarriages and the dates herbicides were sprayed on area forests by two local timber companies - Starker and Williamett - and by the U.S. Forest Service and the Federal Bureau of Land Management. The miscarriages occurred between 1973 and 1977.

The herbicides in question contain dioxins, and are used to destroy vegetation that competes with tree seedlings struggling to grow into commercial timber.

The chemical 2,4.5-T was used by the U.S. military as part of the notorious "Agent Orange" to defoliate jungles in Vietnam to expose enemy troops. But the Defense Department stopped the defoliation program in 1970, after findings that the defoliant caused birth defects in mice and rats.

Hill's letter said that 10 miscarriages "for a population our size (under 1,000)" was unsually high. "Each of us was under the care of a physician at the time of the miscarriage," she wrote, "and none of our doctors could offer any explanation for the miscarriage when it occurred."

She said "it seems more than coincidental that so many of us have miscarried only in the months of March througfh early June" since virtually all of the herbicide spraying occurred between February and April.

Hill said that the only did not occur during that period "happened in October to one of us who lives in an area sprayed in September of that year.

"We are not trying to make rash, unsubstantiated claims," Hill wrote. "But we are interested in seeing if there is a cause-effect relationship. Some of us do know that large acreages were sprayed within a month before our miscarriages."

Last month, in response to the letter, the EPA sent three investigators from its Fort Collins, Colo., research program at Colorado State University to Alsea to interview the women and have them fill out a detailed questionaire.

"It will probably be a number of weeks before EPA officials can decide, based upon the questionaire and other information, whether the herbicides may have contributed to the Alsea women's problems," said an EPA statement yesterday. "If the answer appears to be yes, the agency will consider broader study of the area to see if an abnormal rate of miscarriages exists in the Oregon population."

Most of the women live in mountainous west central Oregon timberland, and were unreachable because they don't have telephones.

But Dr. Norton Kalishman, a health department official who knows all eight, said, "These women, especially Bonnie Hill, have done a wonderful job of trying to come to grips with the probable causes. They did as good a study as most scientists would have done."

Kalishman himself has studied the case, but has no answers yet. "I feel that when you start spraying chemicals and you don't know the effects, you are taking unnecessary chances. Unfortunately, others take a look at it only in te rms of dollars and cents," he said in a telephone interview.