Catherine Arnold picked up a can of beans from the supermarket shelf and pointed to the "Nutrition Facts" label on the back: "Fat, 0 percent," it said, "Cholesterol, 0 percent." She dropped the can into her cart.

"We read the labels all the time," Arnold said, nodding to husband Dave.

"We try not to be obsessed," he added, "but we are."

The Arnolds, a spry retired couple from Falls Church, are worried about heart disease, and closely perusing the nutrition labels has become an essential part of their food shopping.

They are not alone. Study after study shows that a majority of Americans--in some cases huge majorities--routinely use Nutrition Facts to help them decide what to eat.

Food companies have developed thousands of new, reduced-fat products since the labels were mandated by law in 1994, and public enthusiasm for lower-fat diets has held firm.

But in one significant area, the labels have proved a bust. Six years after the Food and Drug Administration predicted the labels would mitigate a variety of health risks, including obesity, Americans are fatter than ever.

A recent Centers for Disease Control and Prevention study classified 17.9 percent of all Americans as obese in 1998--a 6 percentage point increase since 1991. A separate CDC study in 1997 said slightly more than half of all Americans needed to lose weight.

Experts attribute this epidemic partly to causes unrelated to food labeling, including lack of exercise and increased consumption of restaurant meals and take-out food.

"There's so much food around that it's a miracle everybody isn't fat," said Marion Nestle, chair of New York University's Department of Food and Nutrition Studies.

Tufts University nutritionist Jeanne Goldberg acknowledged that the labels' "focus is on fat." Label advocates, she said, "forgot that calories were ever part of the equation."

Disillusionment with low-fat diets may be one reason why the NPD Group, a consumer marketing and research firm based in Port Washington, N.Y., found that fewer shoppers are reading the labels in recent years and a declining number of consumers are regularly eating foods labeled "diet" or "light," or foods low in cholesterol, sodium and sugar.

"The thrust of the nutritional effort is to get people to eat 'healthier' versions of the same things, as opposed to changing to different foods," said NPD Vice President Harry Balzer. "Americans have come to realize it doesn't make any difference. They measure progress by weight, and decide 'I've tried it, and it doesn't work.' "

The 1994 regulations were designed to set government standards for the labeling of processed foods, imposing accuracy and uniformity on conflicting and often misleading information provided by food companies.

The initiative had three objectives: improve consumer knowledge of nutrient content; encourage manufacturers to make healthier products; and decrease rates of cancer, heart disease, osteoporosis, obesity, hypertension and allergic reactions to food.

But then-FDA Administrator David A. Kessler had even higher hopes. Success, he said in a 1993 news conference, "will be measured in whether dietary habits change, in the types of new products that are marketed, in the level of fat consumed, and ultimately in whether Americans lead healthier lives.

The labels debuted in August 1994, listing nutrient percentages, calories and ingredients. They described percentage "daily values" of the Agriculture Department's estimated "average" portion sizes, based on a daily diet of 2,000 calories. In addition, the regulations mandated standards for a variety of product claims, like "light," "diet," "high-fiber," "low-fat" and even "healthy."

By November 1995, an FDA study found that its first objective had been almost effortlessly achieved: 56 percent of consumers used the new labels "often" to check nutrients and compare brands, a 13 percent jump over pre-labeling figures; 48 percent of consumers said they had changed their minds about buying or using a food because of the new label--18 percent higher than in 1990.

Through the years, label use has remained high.

"People look at the label when they buy . . . and stop buying some foods because of what they read," said Bruce Silverglade, director of legal affairs for the consumer advocacy group Center for Science in the Public Interest. "It's an enormous success."

Indeed, in a recent visit to a Giant supermarket in Falls Church, it was impossible to find anyone who did not rely on the labels to some degree, and most were adept at finding exactly the information they wanted, belying the contention of many experts that reading the label doesn't always mean comprehension.

For retirees like the Arnolds, cholesterol is important. Rebecca Stevenson, 25, a young mother from Arlington, checked "only for fat," and said she had lost 40 pounds in one year.

But Otha Easley Sr., 65, said he has "high cholesterol, diabetes, a heart problem, cancer and an ulcer," and needs information on sugar and food additives "all the time."

America Torres, a middle-aged immigrant from Central America, spoke only a few words of English, but said she routinely checks fat and cholesterol for the sweet biscuits she eats with her breakfast: "I have to have them," she said with a grin.

The new regulations also had an immediate and profound effect on food companies, initially reluctant to embrace changes that the FDA estimated would cost them $1.4 billion to $2.3 billion to implement.

"The label produced a more level playing field" by imposing a uniform standard, said Tim Willard, spokesman for the National Food Processors Association, and "they became positive about it."

Between 1995 and 1998 the Chicago-based New Product News reported the introduction of more than 6,500 reduced-fat foods: "There has been a ripple effect," said Silverglade. "Because the label has changed consumer purchasing decisions, the companies have responded with new products."

But many experts said this is not always a good thing. NYU's Nestle agreed that the label succeeded "only too well" in provoking a huge industry effort "to produce food that didn't have any fat."

But the "problem is that calories are calories," Nestle said, and while fat is the biggest source of calories per unit weight in diet, it is not the only one. "We thought that if you cut fat, you automatically cut calories. Wrong."

The anomalies have been noticed at Giant, particularly by younger shoppers: "I look at ingredients to see how much other stuff they've put in it," said Josette Demarest, 39, a computer programmer from Falls Church. "If the first ingredient is water, then you know that the 'low fat' stuff may not be what it seems."

"A lot of times low-fat doesn't mean it's the best," agreed Traci Arnold, 25, a personal assistant to a corporate CEO and no relation to Catherine and Dave Arnold. "It has plenty of sugar to compensate for the fat calories."

Sophisticated shoppers like these, irritated with what Demarest calls "misleading" information, may be why the NPD Group's annual survey of "primary meal preparers" shows a decline in those who "frequently check labels" for "anything I'm trying to avoid" from 60 percent in 1996 to 55 percent in 1999.

A different NPD Group survey of all consumers reflected similar trends. NPD found the number of consumers who regularly ate foods low in cholesterol dropped from 78 percent in 1992 to 52 percent in 1999. For low sugar, the percentage dropped from 51 percent to 31 percent.

"If you take the whole effort, the only thing that Americans have glommed onto is low- or no-fat products," said NPD's Balzer. The survey showed that regular eaters of low-fat foods have held firm at 76 percent for the last four years.

Fat consciousness has translated into some modest reductions in national fat consumption.

This eventually may lead to reduced rates of cancer and heart disease, as Kessler hoped, but Silverglade and others acknowledged that years must pass before such effects can be measured.

That leaves obesity, where labeling appears to have been no help. An October 1999 CDC survey of adult obesity between 1991 and 1998 found the biggest percentage increase (69.9 percent) was among those age 18 to 29 (from 7.1 percent to 12.1 percent).

"Low-fat is misleading," said Arthur Frank, director of George Washington University's Weight Management Program. "The average American may be eating less fat, but the calorie consumption has gone up. There is a mistaken belief that you can eat as much carbohydrate as you want."

Whether labeling is to blame for this ignorance is unclear, for while "the current label screams 'fat,' " noted Tufts' Goldberg, information on calories and sugar is readily accessible on the label.

"I can't tell you what's the most important determinant of long-term health," said Kessler, now dean of Yale's School of Medicine. "It may be a reduction in calories; it may be a reduction in fat."

But it may not matter to label users. "People read it," Kessler said, "and if you want to measure its success, go ahead and propose a regulation today that does away with it."

Staff researcher Lynn Davis contributed to this report.

CAPTION: Catherine and Dave Arnold pause to examine the nutrition label on a can of baked beans as they shop together at a Baileys Crossroads grocery store.

CAPTION: Nutrition Conscious? (This graphic was not available)