ASHINGTON
THURSDAY, I walked into a hardware store in suburban Maryland to buy de-icing crystals in advance of a predicted weekend snowstorm. Lines of customers waiting to pay snaked through the aisles, dozens of men and women with shopping carts full of duct tape and plastic rolls. Needless to say, I left without de-icing compound. I also left thinking, What's the point of this?

Flashing "threat level" warning boxes on newscasts. Police officers with shotguns wandering Times Square, antiaircraft missiles near the Washington Mall. Federal instructions to stockpile water and batteries and obtain plastic and tape for a "safe room." Yet it's far from clear that this security rush will help anyone.

Government cannot, of course, know what will happen or when. During the 1960's, when the menace was missile attack by the Soviet Union, citizens were urged to do both the useful (stock fallout shelters) and the useless (crouch under the desk at school). Officials suggested such things because it was what they were able to think of.

Today, with no sure defense against terrorism in a free society, officials concerned about chemical or biological attack are suggesting the things they are able to think of. But this may only distract attention from the more likely threat of conventional bombs — and the ultimate threat of the atom.

Consider the mania for duct tape. As Kenneth Chang and Judith Miller reported in The New York Times last week, experts view the taped-up room as mainly a psychological benefit. Moreover, many now rushing to buy duct tape may have exaggerated, media-pumped fears of chemical or biological weapons.

If terrorists use chemical weapons, they will probably affect a tiny area at worst, because terrorists would have chemical agents in relatively small amounts. Though any amount of chemical agent might seem ghastly, in actual use chemicals have proved no more deadly, pound for pound, than conventional bombs.

The British and Germans used one ton of chemical weapons per fatality caused during World War I. The 1995 release of the nerve gas sarin in the Tokyo subways by the Aum Shinrikyo sect killed 12 people, fewer than a small, standard bomb might have killed in that crowded, enclosed area. An estimated 5,000 Kurds died in Saddam Hussein's chemical attack on Halabja, Iraq, in 1988, but this involved dozens of fighter-bombers making repeated low passes over the town. It's hard to imagine that terrorists could pull off such a coordinated heavy military maneuver.

A terrorist release of chemical weapons in an American city would probably have effects confined to a few blocks, making any one person's odds of harm far less than a million to one.

Your risk of dying in a car accident while driving to buy duct tape likely exceeds your risk of dying because you lacked duct tape.

Last week, a Washington talk radio host discussed what listeners should do if "a huge cloud of poison gas is drifting over the city." No nation's military has the technical ability to create a huge, lingering gas cloud: in outdoor use, chemical agents are lethal only for a few moments, because the wind quickly dilutes them. Chemical agents are deadly mainly in enclosed circumstances — subways, for example, or in building ventilation systems. The duct-taped room in a home is of little use in such a scenario.

A 1993 study by the Office of Technology Assessment found that one ton of perfectly delivered sarin, used against an unprotected city, could kill as many as 8,000. But the possession by terrorists of a ton of the most deadly gas seems reasonably unlikely, while perfect conditions for a gas attack — no wind, no sun (sunlight breaks down nerve agents), a low-flying plane that no one is shooting at — almost never happen. Even lights winds, the 1993 study projected, would drop the death toll to about 700.

Seven hundred dead would be horrible, but similar to the harm that might be inflicted in a crowded area by one ton of conventional explosives. Because these explosives are about as deadly as chemicals pound for pound, but far easier to obtain and use, terrorists may be more likely to try to blow things up. Almost all recent terrorist attacks around the world have involved conventional explosives.

The image of millions cowering behind plastic sheets as clouds of biological weapons envelop a city owes more to science fiction than reality. The Japanese use of fleas infected with bubonic plague against Chinese cities in World War II was the only successful instance of bioattacks in contemporary warfare. In 1971, "weaponized" smallpox was accidentally released from a Soviet plant; three people died. In 1979, an explosion at another Soviet site released a large quantity of weapons-grade anthrax; 68 people died.

In 1989, workers at an American government laboratory near Washington were accidentally exposed to Ebola, and it was several days before the mistake was discovered; no one died. A coordinated anthrax attack in the fall of 2001 killed five people, a tiny fraction of the number who died of influenza during the time the nation was terrified by the anthrax letters.

None of this means bioweapons are not dangerous. But in actual use, biological agents often harm less than expected, partly for the simple evolutionary reason that people have immune systems that fight pathogens. Also, as overall public health keeps improving, resistance to bioagents continues to increase.

Conceivably, being in a duct-taped room could protect you if a plane dropping anthrax spores were flying over. Smallpox, on the other hand, must be communicated person to person. Those in the immediate area of an outbreak might be harmed, but as soon as word got out, health authorities would isolate the vicinity and stop the spread. By the time you knew to rush to your sealed room, you would either already be infected or the emergency would be over.

Another point skipped in the public debate: smallpox is awful and highly contagious, but with modern treatment usually not fatal. Anthrax doesn't necessarily kill, either, as the nation learned in 2001. Only in movies can mists of mysterious bioagents cause people to drop like stones. In reality, pathogens make people ill; medical workers rush in and save most of the exposed.

If germs merely leave sick people whom doctors may heal, terrorists may favor conventional explosives that are certain to kill.

While government officials now emphasize improbable events involving chemical or biological arms, less is being said about how to be ready for two macabre threats the public is unprepared for: atomic explosion, and the radiological, or "dirty," bomb.

The chance that a crude atomic device will someday detonate on American soil is, by a large margin, the worst terror threat the nation faces. Yet the new Department of Homeland Security has said little about atomic preparedness.

To think the unthinkable, if an atomic device bearing about the yield of the Hiroshima weapon went off outside the White House, people for roughly a mile in each direction might die. But most people in the District of Columbia would survive, while the main effect on Washington's suburbs would be power failures and broken windows. So the majority of people in Washington and its suburbs who would not die would need to know what to do. But do they? Generally not, because there has been scant discussion.

(Here's what to do: Remain indoors at least 24 hours to avoid fallout; remain on ground floors or in the basements of buildings; if you are upwind of the explosion stay put; if downwind, flee by car only if roads are clear since buildings provide better fallout protection than cars.)

Perhaps more likely than an atomic detonation would be a "dirty bomb," in which conventional explosives spread radioactive material. Since this has never been used, effects are hard to project. Most likely, even an extremely large dirty bomb (say, an entire truck converted to one) might kill only those within a city block. Fallout would probably threaten only those a few hundred or thousands of yards downwind.

Yet if people heard on the radio that a dirty bomb had exploded — if they so much as heard the word radiation — panic might set in. In Manhattan or Washington, mass chaos to escape might result in more deaths than the bomb itself.

But is the government explaining to the public how to react if a dirty bomb goes off? (Stay indoors; if upwind do nothing; if downwind, drive away only if roads are clear; take potassium iodide pills to prevent some effects of fallout.) The Department of Homeland Security Web site, for one, has loads of information about anthrax, but offers essentially zero on what to do in the event of radiological explosions.

Increased presence of police and military units in cities may help deter terrorists, and by being more visible and waving bigger weapons, law enforcement is doing what it can think of. But government officials who are advising people to buy plastic sheets create unnecessary anxiety while achieving little beyond helping hardware stores. The advice people need to hear concerns the atomic threat — and why potassium iodide matters more than duct tape.

Gregg Easterbrook, a senior editor of The New Republic, writes regularly about science and government. His latest book is "The Here and Now."