The Honolulu improvement ran counter to the national trends which showed that congestion is now threatening the quality of life for commuters in not only large cities but smaller ones such as El Paso, Texas, and Nashville, Tenn.

Cheryl Soon, city Transportation Services director, said the drop is probably the result of many factors, including a downturn in population and jobs, the opening of the H-3 freeway and widening of H-1, the movement of jobs to the Kapolei area, and efforts to better manage traffic through traffic light synchronization and increased bus service and car pools.

"It's not any one thing, just an incremental series of things we've tried to do to keep up with traffic problems," Soon said.

The nation's most congested area remained Los Angeles, where the average rush-hour traveler endured delays of 136 hours in 2000.

In most mid-size and smaller cities the congestion is getting worse at a faster rate. The growth in congestion was worst in mid-size cities including Cincinnati (59 percent), Austin, Texas (97 percent) and El Paso (110 percent).

"Even cities below half a million haven't been able to slow down the growth of congestion," said Tim Lomas, one of the study's authors.

Honolulu was an exception.

The report shows that the amount of daily vehicle miles traveled on freeways, arterial streets and other roadways had all decreased slightly since 1994, while the number of travel lanes increased slightly. That helped lower the annual cost of congestion in Honolulu from $175 million in 1994 to $155 million in 2000.

Only Salt Lake City; Tampa, Fla.; and New Orleans showed similar decreases.

Even with the improvements, a typical Honolulu driver still spends the equivalent of one full day, 24 hours, stuck in traffic each year, the study said. At its worst, in 1992, the lost time in Honolulu totaled 35 hours.

Soon said she takes such studies with a healthy dose of skepticism. "You've got to watch the trends and methodology, otherwise it doesn't mean much to the guy stuck in traffic tomorrow morning," she said.

For the first time this year methodology of the usually respected annual study have come under some criticism from those who say it does not reflect real-life conditions.

State of Washington Transportation Department spokeswoman Linda Mullen said the study's calculations do not take into consideration congestion relief efforts such as freeway ramp meters, tow-truck patrol programs and traffic monitoring efforts that use freeway surveillance cameras. Such efforts, she said, help move the same number of cars more quickly through the same stretch of road.

The study's authors said such factors will be included in future efforts.