Contaminants in the sediment include toxic metals, industrial compounds, petroleum byproducts and a banned insecticide, all at levels that signal potential cancer risks or other long-term hazards, a News review of government test results shows.

The cleanup plans would involve crews using front-end loaders to scoop up contaminated sediment that Hurricane Katrina floods left in yards, playgrounds and other spots throughout the greater New Orleans area. In some instances, protecting people might require steps less radical than removing soil, such as planting grass to cover contaminated yards.

It's not clear which remedy would apply in which neighborhoods, or how officials will decide. The plans have not yet been completed or made public but were described to The News by several sources familiar with them.

In all cases, however, the task would be complex and huge, with crews covering nearly an entire city and its suburbs while maneuvering around the remaining debris and damaged houses.

Concern about contamination is a major reason the city is allowing residents from some of the most heavily flooded areas only the chance to check on their houses, not stay, said Dr. Kevin U. Stephens, New Orleans' city health director.

Proper removal of the contaminated sediment will ensure that residents won't face undue toxic risks as a result of the floods, Dr. Stephens said.

"If the corps does what it's supposed to do, it should work," he added.

Decisions are still evolving on whatever follow-up testing might occur to check the long-term health of the New Orleans environment after the sediment is gone, said William H. Farland, the Environmental Protection Agency's acting deputy administrator for science.

Local environmentalists complain that they and other members of the public have been shut out of the decisions.

"We've cooperated with the EPA on a great many things over the years," said Wilma Subra, a consulting chemist in New Iberia, La., who is monitoring the hurricanes' environmental impact for the Louisiana Environmental Action Network, an advocacy group. "Now, we're having a hard time getting our calls returned."

EPA spokeswoman Eryn Witcher said the agency is releasing information as quickly as possible, but she and other EPA officials emphasized that they can offer only their best advice. Decisions, such as repopulating New Orleans, are strictly up to city officials, she said.

Information about the possible long-term risks in dozens of New Orleans neighborhoods is crucial as people decide whether to return home and as the city decides where or whether to rebuild.

Despite one widely publicized study that said the Katrina floodwater was no more polluted than typical urban floods, The News' examination of the EPA's tests of flood-deposited sediments reveals long-term health concerns if the contamination were to remain.

Contaminated sediment was always a more serious long-term worry than floodwater, since the water was quickly removed. In September, experts advised the EPA that toxic dust could spread as the sediment dried.

The EPA and Centers for Disease Control and Prevention in Atlanta also warned that comprehensive tests were needed "to identify any widespread contamination or selected hot spots and to ensure the safety of returning inhabitants or for redevelopment."

The EPA's sampling plan, assembled in just a few days in mid-September and quickly reviewed by outside scientific experts, targeted nearly 200 toxic substances in sediment left by the receding water.

Between Sept. 10 and Oct. 1, field crews of up to 100 government and contractor employees took sediment samples from about 300 sites in the New Orleans area. A few additional sediment samples came after Oct. 1.

Workers took only the top layer of soil, trying to avoid any contaminated earth that might have been there before the floods.

The EPA initially told its outside science advisors that it would compare the results to screening levels that the agency's regional office uses to gauge potential long-term health risks in soil at toxic waste sites.

The EPA considers the screening levels to be very conservative, aiming in most cases to make sure that a person exposed to the contaminant for 30 years would have no more than a 1-in-1 million chance of developing cancer as a result.

The levels are not legal limits for contamination in soil. Instead, they guide experts to problems that might need a closer look – like a forest ranger using powerful binoculars to scan the woods for tiny wisps of smoke.

However, the EPA never published a comparison of New Orleans sediment tests to screening levels. Instead, it worked with federal health officials to use other measurements that looked for short-term hazards, a method more likely to protect those most immediately at risk, such as first-responders and rescue workers.

Although it said it found no toxic threats that would be "immediately hazardous to human health," the EPA cautioned people working in the formerly flooded areas, as well as any returning residents, to avoid contact with contaminated sediment.

"We've made it very clear" that the stuff in many yards is dangerous, Dr. Farland said.

But some other experts say the no-contact warning would do people little good under real conditions. "What are they supposed to do, fly over their yards?" asked Ms. Subra.

In an attempt to understand possible long-term health concerns, The News reviewed the EPA test results of every chemical test at every site in Orleans Parish through Oct. 1 and compared them with the EPA's screening levels for residential soil. The raw data was posted on the agency's Web site.

Altogether, the samples contained at least 77 of the nearly 200 chemicals tested for, the review found. Although most were below the screening levels, at least 15 were higher.

Eight of those that were higher than screening levels are known, probable or possible causes of cancer in people. The most widespread was arsenic, a known human carcinogen.

It appeared at virtually every site tested. All but one of the sites that contained the toxic metal had more than the EPA's cancer-risk screening level for arsenic in residential soil, which is 0.39 parts per million.

The highest arsenic level found was 78 parts per million, 200 times the screening level. The vast majority of the sites tested had 10 times the cancer screening level.

Cancer-causing petrochemicals were also widespread, as were toxic components of diesel fuel. About 150 residential test sites had as much diesel as the soil around a leaking underground tank.

A component of creosote called benzo(a)pyrene occurred above the screening level at 100 sites. The highest was 570 times the screening level.

A related chemical, benzo(b)fluoranthene, was above the screening level at 68 sites. The EPA says those two chemicals probably cause cancer, while California state officials say they definitely do.

The banned insecticide dieldrin, used against termites until 1987, showed up at 58 locations. As with arsenic, virtually all the samples had levels higher than the EPA guideline for safe neighborhood soil.

Lead exceeded the EPA guideline at 17 sites. Several others appeared only a few times.

The screening levels alone are not enough to judge the actual long-term risk, EPA officials cautioned. A single site in a neighborhood might not be representative; the surrounding area might actually be higher or lower.

Only a formal risk assessment, a routine procedure in typical toxic waste cleanups, would fully explain how much long-term danger a site poses. But that would require much more information than the EPA has been able to gather, given the extremely tight deadlines and dangerous field conditions that the Katrina disaster imposed.

And the EPA's sampling and cleanup plans addressed only the flood-deposited sediment outdoors. Planning for any cleanup inside homes or other buildings would be a state and local responsibility.

Corps officials expect the sediment removal work to take from 45 to 58 days after plans are approved. The corps is said to be developing four different projects for different areas, but concerns raised by other federal agencies could result in a single plan for the entire flood zone, sources familiar with the planning said.

Corps spokesman Jim Pogue said FEMA would tell the corps to start work after other federal agencies and New Orleans and Louisiana officials agree on the plans.

Environmental groups trying to monitor the cleanup say they want more information on how the government will guarantee people's health for decades, not just weeks.

"We're saying the same thing that we've said since the beginning," said Marylee Orr, executive director of the Louisiana Environmental Action Network. "It's the long-term risk that's the real concern."

Dr. Farland, the EPA scientist, said the agency shares that concern.

"We're not saying ... that there would not be a concern," he said. The main question is "How do we remove the source of any concern?"