A gun kept at home is far more likely to kill those who live there than to protect them from intruders, according to a new study published today.

The study looked at all gunshot deaths over a six-year period in King County, Washington -- including Seattle and surrounding communities -- and found that more than half occurred in the home.

Of firearm deaths in the home, 8 of 10 were suicides and 1 of 10 were criminal homicides, mostly during arguments or fights with family or friends, said the study's author, Dr. Arthur L. Kellermann. Three percent involved accidents.

Only two of 398 firearm deaths in the home -- one-half of 1 percent -- proved to be an intruder shot during attempted entry.

"The great majority of homeowners are probably better off not having a protective firearm," said Kellermann, who published his findings in this week's issue of the New England Journal of Medicine. "If they keep guns, they should keep them securely stored, with ammunition stored separately, if at all, in the home."

Kellermann and other researchers said that despite widespread political debate over gun control, there had been little unbiased scientific research on injuries and deaths from firearms. He said this appeared to be the first major study of firearms that focused on mortality in the home, but emphasized that it was "directed more toward individual choice than a national firearms policy" and "doesn't address gun control per se."

The study noted that there are about 120 million privately owned weapons in America, with about half of all homes containing one or more firearms. It cited previous studies showing that although most gun owners have them primarily for hunting or sport, three-fourths say they keep them at least partly for protection and one-fifth identify "self-defense at home" as their most important reason for owning a gun.

"The problem of firearms is not just a ghetto problem or a street problem. It's Ma and Pa America in their living room getting in a fight and one of them picking up the family firearm. It's your next-door neighbor in a fit of deep depression and possibly intoxicated, picking up a gun and ending his life," said Kellermann. "It's a problem we all share as Americans."

Kellermann is an assistant professor at the University of Tennessee in Memphis, and director of the Emergency Department at the Regional Medical Center there. The study was conducted while he was on a private research fellowship at the University of Washington, in conjunction with Dr. Donald T. Reay of the King County Medical Examiner's Office.

It found that from 1978 through 1983, 743 deaths from gunshot were investigated. Firearms were involved in almost half of all homicides and of all suicides in King County, levels somewhat lower than the national average. King County, with a population of 1.3 million, is largely urban and the population is mostly white.

The study's main findings:

*Of the firearm deaths, two-thirds occurred inside a house or dwelling, and 398 -- 54 percent -- in the home where the firearm involved was kept. Handguns were used in 70 percent of the deaths. Of the 398 deaths, 333 involved suicide, 41 were called criminal homicides, 12 were due to accidents and three were unknown.

*Of deaths in the home, nine were classified as "self-protection," with two the shooting of burglars by residents and seven involving homicides committed in self-defense during fights. Kellermann calculated there were "43 suicides, criminal homicides, or accidental gunshot deaths involving a gun kept in the home for every case of homicide for self-protection."

*Most of the victims were male and alcohol was often a factor. Of suicide victims, 80 percent were male, with one-third found positive on a blood alcohol test.

*Of 65 nonsuicidal deaths, 55 percent involved residents, including relatives, spouses and roommates, while one-third were friends or acquaintances. More than 80 percent of the homicides occurred during arguments.

Kellermann noted that the study had no way of documenting cases in which intruders are wounded or frightened away by use of a gun.

Paul Blackman, research coordinator for the National Rifle Association, said he had not seen the Kellermann study. But he derided it as "meaningless" and "simplistic" in focusing exclusively on deaths as well as ignoring "all the beneficial purposes" of firearms.