The routine act of driving has become the riskiest thing most Americans do, producing more horrific body counts than any modern war or terrorist act.

More than 100 people perish every day on America's killer roads.

And the most deadly stretch of road in the country? Interstate 15 between Las Vegas and Los Angeles, according to Scripps Howard News Service's "Killer Roads" national reporting project.

While the carnage nationwide has fallen in recent years, 37,261 individuals died in vehicular accidents in 2008 — that's still more than 10 times the number who died in the attacks on Sept. 11, 2001.

Most of the traffic fatalities can be attributed to excessive speed, alcohol-impairment or failure to wear seat belts. Drivers distracted while texting, eating or using their cell phones are also a growing concern.

But often the roads themselves are to blame.

"We know there are sections of highways that are more dangerous, since they have more fatalities," said former Federal Highway Administrator Kenneth Wykle. "Those locations need rigorous accident-investigation analysis. What are the contributing factors? There must be a cause."

As part of "Killer Roads," Scripps counted the number of deaths on every road in America and ranked the worst roads in each county.

Individual reports on the nation's 3,100 counties — including maps plotting the most recent fatalities color-coded by risk factor — can be seen at www.scrippsnews.com/killerroads.

Using data provided by the U.S. Department of Transportation, Scripps analyzed 562,712 fatal accidents from 1994 to 2008 that claimed 627,433 lives.

Sometimes, the deadliest roads seem disarmingly safe — a small country lane winding gently through rolling hills or a perfectly straight superhighway stretching across a vast desert landscape.

The study found that the single most deadly road is the 181-mile segment of Interstate 15 that passes through San Bernardino County, Calif., and connects Los Angeles to Las Vegas. During the 15-year period of the study, 1,069 people perished in 834 accidents. That's more than double the death toll from any other road in any other county.

"It isn't really the road's fault. I've driven it many, many times," said Holly Vogel of the California Department of Transportation. "But it's a straight route out in the middle of the desert where you can see for miles and miles. So people start speeding up and stop paying attention."

According to federal files, more than half of the people who died on that section of I-15 were not wearing their seat belts. About a third of the accidents involved speeding, and nearly a quarter involved at least one driver who'd been drinking.

The roads with the highest death counts are predominately heavily used interstate highways like I-15. But the Scripps study also found that small country roads have become some of the worst killing grounds. Thousands of motorists died on roads that are reported in federal records to have "no name."

Nearly 245,000 accidents occurred along country roads, small municipal streets and other minor routes, killing more than 265,000 people. Because most of these roads are lightly traveled, their actual death rates per miles driven can be much higher even than along Interstate 15.

"A disproportionate number of people are losing their lives along rural, two-lane roads," said former U.S. Transportation Secretary Mary E. Peters. "This is something absolutely that we have to tackle. More guardrails can help, and we can widen shoulders so we have fewer vehicle flip-overs. And there are long-term solutions like straightening out the curves."

Motorists also appear to drive differently on different kinds of roads. The Scripps study found that 24 percent of all fatal accidents along interstates involved drinking. But drinking was reported for fatal crashes 31 percent of the time on state roads and 39 percent of the time on county roads.

"People may feel more comfortable drinking and driving in rural areas, thinking that they are not as likely to get caught as on major roads," concluded Lee Munnich, director of the Center for Excellence in Rural Safety at the University of Minnesota. "The other problem is that when people are in crashes, they are less likely to be wearing a seat belt in rural areas. They may feel they won't be observed not wearing seat belts."

Munnich's group has created a Web site at SafeRoadMaps.org and has plotted thousands of crash sites and identified dozens of "hot spots" where accidents appear to occur at suspiciously high levels.

Can something be done to re-engineer hundreds of thousands of miles of country roads?

"The answer is yes," Munnich said. "There is only so much we can do with rural roads because there are so many. But we found things like roundabouts — traffic circles — are safer than intersections. And rumble stripping (along the shoulders) helps warn people they are leaving the roadway."

The study found enormous geographic differences in the reported rates of risky behavior. Speeding was reported in only 9 percent of the fatal accidents in New Jersey but cited 44 percent of the time in South Carolina.

"One of the factors may be the roads themselves," said Wykle. "New Jersey has more national-highway-system roads like interstates, turnpikes and four-lane roads. South Carolina is more rural and, so, the conditions are different. You can't go as fast (safely) on those roads. A little error by a driver is more likely to cause an accident in South Carolina than in New Jersey."

The study also found that, while most accidents still involve well-known risk factors, the percentage of fatal accidents without an obvious cause has been rising. About 81 percent of all fatal accidents in 1994 were associated with drinking, speeding or failure to wear seat belts. By 2008, the figure had fallen to 73 percent.

Federal highway safety officials are increasingly concerned that distracted driving is a growing threat, as use of cell phones and texting devices behind the wheel has increased dramatically in recent years.

Aggressive campaigns against traditional risky driving behavior have been widely credited with lowering annual traffic fatalities from 43,000 in 2002 to slightly more than 37,000 in 2008.

"But," Peters said, "it's still a national tragedy because far too many people are dying."

e-mail: [email protected]