Choosing to use biotechnology has been something of a no-brainer in U.S. agriculture the last few years. Plants that make their own insecticides mean lower costs, higher yields, bigger profits. Domestic public acceptance is high and the regulatory climate has been supportive.

The only dark cloud, a small and distant one, has been a spate of Luddite-flavored protest overseas. But now bigger clouds are gathering, and quickly. Large overseas consumers -- Japanese breweries, Mexican tortilla factories, Britain's biggest grocery chain -- are rejecting genetically modified food products because of possible risks to consumers. And the European Union, after accepting 18 such products for import since 1990, has stopped approving any new ones.

Suddenly it seems possible that old-fashioned corn, cotton and soybeans may bring higher prices than their engineered competitors, at least in export markets. And suddenly the promoters of biotech foods are being reminded in the plainest terms that customers are still king. All the safety assurances in the world are worth nothing if the buyers aren't persuaded.

This point is worth special attention at the leadership level of the biotech companies, primarily Monsanto and Novartis, and of the federal regulatory agencies, chiefly the Food and Drug Administration, Department of Agriculture and Environmental Protection Agency. All have been partners in a safety-assurance effort that blithely assumes the new foods safe unless proven otherwise, and arrogantly resists labeling goods with biotech content on the ground that consumers will irrationally reject out of prejudice.

It is true, as the advocates say, that no harm to humans has been scientifically linked to any of the new food products -- which have become so ubiquitous, via the oils and starches and sweeteners essential to food processing, that only the most rigid organic-food adherents are likely to avoid eating them. But it is also true that neither the FDA nor other agencies has done much long-term, independent testing of the products, relying instead on industry assurances that they are not significantly different from their nonengineered counterparts.

The possibility of risk to consumers has been underlined by the notorious accident in which allergenic proteins were spliced from a Brazil nut into a soybean. And the prospects of environmental harm -- a separate but equally significant issue -- are suggested by studies showing that biotech crops can kill monarch butterflies and other beneficial insects, and may be giving rise to pesticide-resistant bugs and weeds, as well as microbes that can't be killed with antibiotics.

Such developments have led the USDA, belatedly, to intensify its monitoring of environmental impact from the new products and to review its process for approving them. The FDA and EPA, meanwhile, are still working on the detailed regulatory policies that were promised five years ago. No wonder the industry's blithe assurances of no proven harm are getting a skeptical reaction in other parts of the world, especially in Western Europe, which lately has had more than its share of genuine food scares. The new products may indeed prove to be risk-free, but the demands for that proof -- and for appropriate safeguards in the interim -- are rapidly getting larger, not smaller. By moving in the opposite direction, the market is making this point with unmistakable clarity.

** NOTICE: In accordance with Title 17 U.S.C. Section 107, this material is distributed for research and educational purposes only. **