(FORTUNE Magazine) – The cheese course at Manhattan's Picholine restaurant is a theatrical experience. The cart glides toward you like a gondola on wheels, arousing looks of envy as it passes. With appropriate aplomb, maitre fromager Max McCalman presents you with "The Cheeses of Picholine," a guide to the geographic origins, type of milk (cow, goat, or sheep), and flavor characteristics of the 70-odd varieties.

What the guide doesn't say is that several cheeses on the list are illegal; according to the Food and Drug Administration, they can't be imported or sold in the U.S. Moreover, the FDA is considering stricter guidelines that would ban outright many more cheeses on Picholine's list.

Even if you can't tell the difference between a Portuguese Azeitao and a French Fourme d'Ambert, your favorite cheeses may still be at risk. In fact, the Parmagiano-Reggiano you grate over pasta and the Gruyere you melt for fondue would both be illegal under the new standards the FDA is contemplating.

For more than 50 years the FDA has held the belief that pathogenic bacteria--including salmonella, E. coli 0157:H7, and Listeria monocytogenes--are present in raw milk and can contaminate cheeses made from it. As a result, the agency requires that cheeses manufactured from raw milk be aged for 60 days at a temperature of at least 35 degrees Fahrenheit. Last year, in a climate of increasing fear over the spread of food-borne pathogens in beef and under pressure from President Clinton's Food Safety Initiative, the FDA began to review an outright ban on raw-milk cheeses. "We became aware of some reports in the scientific literature that suggested certain human pathogens likely to be in raw milk can survive the 60-day aging process," says Jack Mowbray, a regulatory policy analyst at the FDA. Mowbray also cites several known outbreaks from raw-milk cheeses, including a 1984 outbreak of salmonella in Canada in which 2,700 people were exposed--and one died--after ingesting cheddar cheese made from raw milk.

Mowbray emphasizes that no conclusions have been reached; the FDA is simply being cautious. To that end, the FDA authorized the National Center of Food Safety and Technology (NCFST) to conduct tests by inoculating raw-milk cheddar with a strain of E. coli 0157:H7 isolated from food that was responsible for a prior outbreak. "We wanted to see what could happen in a worst-case scenario," says Mowbray.

The FDA's approach seemed prudent, until FORTUNE discovered that the cheddar in the Canadian outbreak wasn't made from raw milk at all. Some was pasteurized and some heat treated (a partial pasteurization). In fact, pasteurized milk was the culprit in the majority of dairy-related outbreaks, including the largest food-borne epidemic ever reported to the Centers for Disease Control--a 1985 salmonella outbreak from pasteurized low-fat milk that led to 2,777 hospitalizations and 14 deaths. Even in cases directly implicating raw-milk cheese, the argument could be made that contamination occurred during manufacturing and was unrelated to the presence of raw milk.

"The biggest risk from cheese comes from post-pasteurization contamination," says Dun Gifford, head of Oldways, a nonprofit dedicated to preserving natural foods. "It's like saying if there's a salmonella outbreak in chicken, you should outlaw all chicken," adds Peter Kindel, maitre fromager of Artisanal, a new restaurant and cheese shop from the owners of Picholine. "The answer, as with chicken, should be better regulation."

If this is indeed the case--and the FDA knows it--it would explain why restaurateurs currently find it easy to obtain illegal cheese. "We get lots of other raw-milk cheeses aged less than 60 days," says Kindel. "Either the FDA doesn't know what we're bringing in or there's something else going on."

Certainly there's food here for conspiracy theorists. The organization conducting the tests for the FDA, the NCFST, is composed of food technologists and industry types--the very groups that have the most to lose from the small, artisanal cheesemakers forsaking pasteurization and other industrial processes. And in Europe a similar organization, the Institute of Food Science and Technology, has blamed raw milk for a number of outbreaks in which the evidence pointed to the possibility of contaminated manufacturing conditions. Moreover, Mowbray acknowledges that a critical gap exists in the NCFST's research: No one knows what the worst-case E. coli scenario actually is. "We don't have a lot of hard information on which to base our assumptions," he says. "We discussed them with industry and aca-demia, and made the best guess we could."

The FDA doesn't expect to make any determinations for a year or two. Until then, we can only hope that it will decide to put a warning label on all cheese--pasteurized and raw--as it does for alcohol and tobacco, instead of lowering the curtain on that luscious Fourme d'Ambert.

FEEDBACK: personal@fortunemail.com