(FORTUNE Magazine) – A SOBERING THOUGHT strikes you upon entering the National Center for Food Protection and Defense: This place is not very well defended. Located on the second floor of a red-brick building at the University of Minnesota, the $15 million center--launched in July 2004 by the Department of Homeland Security--doesn't look much different from, say, the average college anthropology department. There are no guards at the entrance, no retinal scans required. On a recent sultry morning, the front door isn't even locked. "It only locks after business hours," associate director Shaun Kennedy admits sheepishly as he leads a visitor upstairs.

There's little need for high security--this is the NCFPD's administrative office, and its research projects are scattered elsewhere on campus and around the country. But the center's accessibility serves as a fitting, and disturbing, symbol of the vulnerability of our nation's $1.25 trillion food industry, which the NCFPD is working feverishly to address.

When Secretary of Health and Human Services Tommy Thompson stepped down last December, he made this startling comment: "I, for the life of me, cannot understand why the terrorists have not attacked our food supply, because it is so easy to do." Although Thompson was roundly criticized for his bluntness--some said he was rolling out a red carpet for terrorists--nobody has questioned the accuracy of his statement, one that he reiterated to FORTUNE. "I still put this front and center," he says.

Simply put, with an estimated 2.1 million farms, 900,000 restaurants, 115,000 food-processing plants, and 34,000 supermarkets, the vast, fast-moving U.S. food supply makes an inviting target for terrorism. Two Stanford researchers caused a mild uproar in late June with a study concluding that a third of an ounce of botulinum toxin poured into a milk tanker truck could cause hundreds of thousands of deaths.

The Minnesota center's task is to keep such a scenario purely theoretical. Here, and in partnering universities, 120 specialists in everything from microbiology to economics are working on 26 food-related security projects that span all facets of the food supply chain--production, storage, processing, distribution, and retail. Or as director Frank Busta puts it, "from field to flatulence."

Busta, a garrulous fellow with a raspy voice, has nearly 40 years' experience in food science and technology. When the Food and Drug Administration in 2003 asked an industry group that he led to review aspects of food safety, "it was one of those 'Oh, crap' situations," he recalls. "I can see vulnerabilities in an awful lot of places."

A day of rapid-fire discussions with some of the research teams is by turns enlightening, terrifying, and encouraging. Our first chat is with Amy Wong at the University of Wisconsin at Madison, who is working on a portable gas-stream technology that zaps bacteria from surfaces like stainless steel--useful if a food-processing plant ever needed some heavy-duty decontamination. On the phone from her lab, Wong explains that her "cold plasma" approach doesn't leave the mess that chemical sanitizers do, nor does it corrode the surfaces being cleaned. (Ecolab, a big developer of sanitation products for the food industry, says it could be interested in such a product. But "if we ever made money on this, that would not be a good thing," says Dr. Bruce Cords, Ecolab's VP of food safety and public health.)

Next up is microbiologist extraordinaire Eric Johnson, also at UW-M, who's working on an instrument to detect botulinum. Johnson was part of the team that created the first batch of the wrinkle-remover Botox, and he has written more about the deadly bacterial toxin than Thoreau wrote about the woods. His device detects even scant amounts of the toxin and thus could be put into a processing line for, say, milk, shutting down production before catastrophe struck in the form of a botulism outbreak. The sensor hasn't been field-tested yet, but Johnson says there's a lot of interest, especially in the wake of the brouhaha set off by the Stanford researchers.

The highlight of the day is a test drive of an interactive "food war game" software program that can simulate, with alarming precision, the hour-by-hour devastating effects of a botulism outbreak at an ice-cream plant or an anthrax attack on lettuce. (If you're guessing that both botulinum and milk are major concerns among those in the food industry, you're right. Milk comes from innumerable small farms, making it hard to monitor; botulinum is among the most toxic pathogens on earth.) The software lets you choose how the attack will play out: the food source, the percentage of product contaminated, the public health agencies' response time.

Jeff Sholl, a ruddy-faced former student of Busta's and the software's creator, chooses an ice-cream scenario. He sets his parameters and clicks the mouse. A map of the U.S. appears--it's reminiscent of the 1983 movie WarGames--and all hell breaks loose onscreen as the tainted ice cream moves swiftly from distribution centers to retailers to American homes. The first illness comes on the third day. Five days after the contamination, the deadly product is finally identified, but by then it's too late: More than71,000 people have consumed the tainted ice cream, 15,000 are ill, and 7,500 are in the hospital. "The ice-cream folks said this was plausible," Sholl remarks casually, as he ponders what action to take in response to the outbreak. The final tally: 67,122 dead and a $314.3 million blow to the economy.

After a light lunch (quick diet tip: To shed pounds fast, spend your days talking about tainted food), we take a short drive to Minnesota's St. Paul campus, where we meet Vivek Kapur, a native of Bangalore, India, who's heading work on a portable detection device that first responders could use at the site of an emergency to detect pathogens quickly. "If successful, it would take a three-day approach and make it under an hour," says NCFPD's Kennedy. Companies like Cargill have expressed interest in using the device to verify that a processing plant for luncheon meat, say, has not been tainted with listeria.

As impressive as the projects and scientists are, unanswered questions hang over every presentation: Will food companies pay for any of this fancy technology? Will they pass the costs along to consumers? Food manufacturers and retailers work on razor-thin margins, and security spending "doesn't make an extra ounce of flour," says Ken Schuman, a plant manager for Star of the West Milling Co. in Ligonier, Ind.

Jean Kinsey, co-director of UM's Food Industry Center, has the unenviable task of answering those questions. Her team of economists has devised surveys to measure companies' current spending on food security and whether consumers see value in such spending. Her early interviews with a range of companies have been disheartening: "We're not going to spend any money on this unless we have to," she recalls them saying.

And they don't have to. Beyond the Bioterrorism Act of 2002, which mostly requires food companies to keep careful records and make them available to authorities, there is little in the way of new government-mandated food-security regulations. The onus is on the private sector to raise the bar, but budgets are tight. Says Mark Cheviron, security director at Archer Daniels Midland: "We don't want to overspend on security-related things that don't make sense. A grain elevator does not have same threat level as a nuke plant." Indeed, Burger King, an early partner of the Minnesota center, has backed away from the program "until we see some real returns," says security chief Steve Grover.

Burger King can have it their way, but will others follow? "Can we afford one terrorist act, one incident of sabotage?" asks Schuman. "No, we can't." And that's why Busta's brain trust can't afford to rest. The private sector may be skeptical, but at least Secretary Thompson has put the issue on the front burner

FEEDBACK mboyle@fortunemail.com