The Peacetime Perils of Soldiering. Driving a car on liberty may be the most dangerous task a soldier, sailor, airman, or marine performs. MILITARY SAFETY

HELICOPTERS fly into power lines at night. A battleship turret explodes. B-1 bombers belly-flop into the desert, while armored personnel carriers turn over and cratering mines explode prematurely. Even in peacetime, being in the United States military can be dangerous. Mixing complex weapons, hard training, and high-spirited youth often leads to accidents - sometimes fatal ones.

But focusing on tragedies such as the USS Iowa explosion gives a distorted picture of armed-service safety, Defense Department officials say. Certain jobs, such as carrier aviator, can be very dangerous, but safety statistics show that on the whole:

Members of the military are most at risk when driving their own autos.

Peacetime military fatality rates are roughly equal to those in mining or farming.

Take the Navy as an example. Besides the Iowa explosion, the Navy has in recent years suffered disasters ranging from an Exocet missile hit on the USS Stark to the downing of an attack jet in a raid on Syrian gun emplacements.

From 1979 to the end of 1988, the Navy had a total of 5,027 fatalities. Yet 2,093 of those were caused not by gunfire or plane crashes or even training accidents, but by motor vehicles. A total of 65 deaths were caused by ``hostile action'' such as the attack on the Stark, according to service records. Homicides accounted for 241 of the Navy's victims.

The Army, largest of the armed services in numbers, has an even higher percentage of deaths occurring in autos. Of the service's 391 fatalities in 1988, two-thirds occurred in private motor vehicles.

``The typical scenario is an accident involving a specialist 4th class, late Friday night, maybe in the rain,'' says an Army safety official.

In recent years the Army has also been getting safer, with fatalities steadily dropping from the 439 that occurred in 1985.

Aircraft crashes are the No. 2 Army cause of fatalities, with 39 last year. Most of those occurred in two helicopter incidents, one a midair collision at Fort Campbell, Ky., and the other the crash of a CH-47 Chinook in a Texas field exercise.

On average only 2 to 7 percent of Army fatalities involve tanks, personnel carriers, and other combat vehicles. In one such incident last year, a tank gunner became disoriented during a night exercise, tracked his turret too far to the left, and fired a practice shell into a companion tank nearby.

The road is also the most dangerous place for the Air Force, which had 48 deaths from aircraft crashes last year and 183 ground fatalities, most in autos.

The ground fatalities figure is the lowest the Air Force has ever recorded. Of the auto accident victims, ``more than half weren't using seat belts or crash helmets,'' says an Air Force spokeswoman. ``Alcohol was a factor in one-third of the mishaps.''

Comparing the safety record of the military to civilian industry is difficult. Statistics are hard to come by, and often are not strictly comparable. Overall statistics do not measure the difference in risk run between a pilot, and a mechanic or a clerk.

If fatalities are used as the measure, a rough comparison seems to indicate that military service is more dangerous than the average civilian industry - but no more or only slightly more dangerous than traditionally hazardous civilian occupations such as mining, agriculture, and construction.

The Army's fatality rate per 100,000 active-duty troops was about 50.1 last year, for instance. The Air Force's was 38.88.

According to the National Safety Council, the average fatality rate for all industry in 1987, the most recent year for which full figures are available, was only 10 per 100,000 workers. But for agriculture it was 49 per 100,000. For mining the figure was 38; for construction, 35. (Typically, one-third of work fatalities are auto related, according to Safety Council figures.)

Another way of comparing military safety to civilian is to examine the military activity widely thought to be the most dangerous - flight. Much military flight is low-level, with aircraft that are sports cars to the civilian world's economy models. Military pilots must often operate in horrible weather.

Last year, the Air Force had a major accident rate of 1.63 every 100,000 flying hours. The Army, which flies mostly helicopters, had a major accident rate of 1.84 per 100,000 hours in the air. The Navy, which flies mostly off carriers, had a comparable rate of 2.1.

These figures mean flying in the military is far more dangerous than taking civilian airliners. In 1987, the airline industry had 0.031 major accidents per 100,000 hours.

But horror stories notwithstanding, major airlines are an extremely safe way to travel. General civilian aviation - using small planes such as Pipers and Cessnas - has a safety record only somewhat better than the military, with a fatal accident rate of 1.45 in 1987, according to the National Safety Council.

Within the overall military, flight figures show some interesting disparities. At first glance, it would seem that the Navy needs to work on its safety procedures, having as it does a worse record than its brother services. But the Navy's record is worse partly because it has a higher percentage of fighter jets, the most dangerous kind of plane, than the other services.

Of US fighters, the F-16 had much the worst safety record last year, with a 6.80 major accident rate per 100,000 hours. By comparison, the F-14 figure was 5.48; the F-15, safest of tactical jets, had a 0.50 rate, according to the Air Force.

As the recent downing of a B-1 by a pelican shows, some military crashes are more ignominious than others. At jet speeds, pilots must see a bird a half mile away to avoid it if on a collision course; bird strikes have cost the Air Force four planes in the last six years.

But military plane crashes are only a small fraction of what they were 35 years ago, when the Navy's major accident rate alone was 10 times higher than it is now. And even if a plane goes down, the advent of ejection seats now means pilots have a good chance of getting out alive. Though the Germans experimented with ejection seats in World War II, they did not become standard issue in the US until the early '50s. Today, about nine out of 10 pilots who ``punch out'' survive. And pilots who eject over water, and who land unconscious, even have their life vests automatically inflated, thanks to a sensing device that became standard issue in 1980.

But even as technology brings more sophisticated safety devices, it can cause new hazards. An Army safety bulletin warns that soda machines have now become so large and heavy that they pose a real threat to anyone who becomes annoyed and rocks them. Since 1984 the Army Safety Center has reported 41 soft-drink machine accidents - including three fatalities.