PHOENIX -- The little brown creatures were squirming and fighting -- that much could be seen with the naked eye. But it took a special microscope to understand what made them unique. They glowed bright green.

The fluorescent larvae are genetically engineered versions of a major crop pest, the pink bollworm. Their green glow comes from a jellyfish gene, spliced into bollworm DNA as a test to see if the animals can survive, and whether it will affect their behavior. But the ultimate goal of the research is far more ambitious -- to use genetic engineering to control an insect that loves to eat the seeds of the best cotton grown in the Southwest.

The effort has progressed far enough that scientists have applied for permission to release some of the bollworms into outdoor cages. If approved, the experiment would be the first planned release of genetically engineered insects outside the high-security laboratories where they are being bred.

The pink bollworm project represents the furthest advance in an explosion of research worldwide into how insects can be genetically modified to benefit humanity. Major projects are underway to create mosquitoes that can control malaria, honeybees that can better fend off disease, silkworms that will produce more silk, and dozens of other normally destructive crop pests that will exterminate their wild brethren.

But those squirming pink bollworms have raised scientific and environmental concerns about the possible unintended consequences of releasing billions of modified insects into the world. Critics say that regulations to control the engineering and release of insects are inadequate and that too little is known about long-term risks.

Genetically engineered corn, canola and other crops have accidentally cross-pollinated and mixed with conventional crops, demonstrating how difficult it can be to control engineered life forms. Critics want more assurances that modified insects won't similarly spread their genes -- raising the possibility that unplanned crossbreeding could kill off other useful insects by making them weak or infertile, or could create super-mosquitoes or crop pests that are far more harmful and difficult to control than today's varieties.

Robert Staten, director of the Agriculture Department's Animal and Plant Inspective Service lab here, is leading the pink bollworm effort, and he knows his work is being scrutinized.

Many see his request to test the larvae outside the laboratory as an important step in the evolution of biotechnology, both its technical abilities and the public's willingness to accept it. But Staten, who grew up in a cotton growing area of New Mexico, sees it more simply as a new front in the war against a costly nuisance.

"There's nothing like this anywhere else in the world," he said, as he studied the glowing engineered pink bollworms through a fluorescent microscope. "They're kind of nasty beasts. 'Pinkie' is a major pest and, left uncontrolled, will strip a field of pickable cotton. We're trying to do something about that."

These genetically modified insects are growing inside a sterile lab in an industrial section of Phoenix, with tight air circulation and locking controls to meet the same kind of biosecurity standards imposed on dangerous viruses. But Staten hopes that by the time the cotton growing season is underway early this summer, a permit for his experiment will be granted and 2,350 engineered bollworm moths will be flying inside screened cages on a well-protected USDA field plot nearby.

What researchers are trying to create with the bollworms is the world's first genetically engineered pest control system. For seven years, they have been working to produce a modified bollworm that would impregnate wild populations with a gene that would stop the development of all fertilized eggs.

If the technology works, it could provide a cheap and effective way to control one of the most damaging cotton pests around. Because cotton farmers have used large amounts of pesticide to control the pink bollworm, the genetically modified moths could be a boon to the environment.

"This could potentially be a major breakthrough in pest control and reducing the risks of pesticide use," said Robert Rose of the USDA, which will review the release permit as part of its regulation of plant pests.

Rose said his agency is conducting an environmental assessment of the planned caged release. The results will be published in the Federal Register, and the public will be asked to comment and raise concerns -- probably next month -- before the experiment is approved or denied.

To ensure that the bollworms do not escape their outdoor cages and mate, females' wings will be clipped and both sexes will be sterilized through irradiation. Researchers and regulators say they don't want any problems with the bollworm release to jeopardize future work modifying insects.

But the likelihood that a successful bollworm effort would lead to the development of many other engineered insects is exactly what has alarmed biotechnology critics. They worry about unknown and unforeseen consequences from mass releases of engineered insects -- newly created entities that cannot be brought back once widely released. The bollworm project may be solid and useful itself, they say, but it could be a harbinger of riskier efforts.

"I'm troubled about the lack of rigor in assessing the environmental risks involved with these genetically modified insects," said Jane Rissler of the Union of Concerned Scientists. "They're impossible to control in the wild because they're so small and mobile and they reproduce so fast. They're also so important in the whole web of life, and we worry that something irreversible might be happening here without all the scholarly and regulatory work that's required."

That concern is shared by University of Florida researcher Marjorie Hoy, who speaks from experience because she conducted the nation's first field test of a genetically modified arthropod (a beneficial mite) several years ago. The experiment required a long and complicated permitting process and resulted in the modified mites losing their engineered characteristics in the field.

But Hoy says that the Agriculture Department's regulation of genetically modified insects is insufficient and that there are "gaps" in the system.

"I think there is promise in genetic engineering of insects for pest management purposes, but we need to go slowly and incrementally and not oversell the technology," she said. "We've been surprised in the past with new technologies like DDT, which was once considered an important solution to pest control. The possible benefits are important, but let's make sure we examine all the risks before any field releases."

William Brown, science adviser to the Interior Department under President Bill Clinton and now with the National Audubon Society, also believes adequate regulation is not yet in place. "This may well be an appropriate use of genetic engineering for biological control, but I don't think there is any statute that really covers it properly when it comes to regulation."

The main beneficiaries of genetically engineered pink bollworms would be the cotton growers of California, Arizona and New Mexico, who have been struggling with the pest since it arrived (probably from Egypt or India, via Mexico) in the early 1900s. In fact, the California Cotton Pest Control Board, an industry-funded panel that advises the state agriculture secretary, has been funding much of the research and development of engineered pink bollworms.

Pest control and biotechnology companies have shown little interest in genetically modified insects -- partly because of the technical problems and public skittishness, and partly because companies (especially Monsanto Co.) have developed genetically engineered cotton that controls pink bollworms. But growers say treatment remains costly and scientists fear that resistance to the engineered toxin will gradually grow and make the product less useful.

"We look at the [genetically engineered] bollworm as our great rescuer on the horizon," said Wallace Shropshire, a cotton grower and chairman of the pest control board. "Because if the technology works as we hope, then we could be talking not just control of pinkie, but total eradication in some of our areas."

The cotton pest control board already has worked with Staten to produce millions of sterile bollworms that are also being used to control the pest in central California. But those pink bollworms, which are made sterile through irradiation, are expensive to produce and not very effective because they are weakened. Researchers say the genetically engineered bollworms would be cheaper and much stronger. Because there is a modern facility in Phoenix to breed the irradiated bollworms, the same plant could also be used to grow billions of engineered insects.

The research on producing modified pink bollworms has been done largely at the University of California at Riverside under the guidance of Thomas A. Miller. That's where the foreign genes were inserted into bollworm eggs and the first stable strains of modified bollworms were produced. The work in Phoenix is continuing and strengthening the line.

According to Staten, pink bollworms are ideally suited to be the first genetically engineered insects to be released, because the risk is low. Most important, he said, the insect is not native to the United States and has no close relatives here. That means the risk of gene transfer to related insects, a major concern of environmentalists, would be virtually eliminated.

The researchers involved with the project are aware of the public discomfort about their work, and they say they are taking great precautions to eliminate risks from their experiments.

"We are prepared to address every fear and concern as it comes up because we really care what the public thinks of this project," said John J. Peloquin, part of the team at UC-Riverside. "We're not doing this because we don't care what happens to our environment, but because we do care so deeply and want to produce something of real public benefit."

** NOTICE: In accordance with Title 17 U.S.C. Section 107, this material is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. **