Las Vegas' effluent is diluted as it flows downstream, and some of the compounds in it are degraded by sunlight, destroyed by microbes, or bound up in sediment. Still, monitoring in 2006 showed that water entering San Diego's municipal system contained, before drinking-water treatment, small but measurable quantities of ibuprofen, the insect repellent DEET, and the anti-anxiety drug meprobamate.

 "We can have a lot more monitoring and control if we oversee our own reclamation than if we're relying on a river with a billion gallons of recharge from other sources every day," says Bruce Reznik. "It's better to be drinking our own 'toilet to tap' water than someone else's. I'm pretty confident that this would be a whole lot safer than what we're getting now."

 In Windhoek, Namibia, water from a wastewater treatment plant is piped right back into the drinking-water system. NASA is developing advanced recycling technology that will directly convert astronauts' urine into clean drinking water. Such reuse systems are what a South African pioneer in water reclamation, Lucas van Vuuren, was thinking of when he said, "Water should be judged not by its history, but by its quality." Sufficient treatment, he meant, assures that any water can be reused. Windhoek is achingly dry and almost 500 miles from the nearest perennial river. It costs NASA about $40,000 to send a gallon of clean water up to the International Space Station. In those situations it makes a lot of sense to clean - carefully - and reuse wastewater.

 Van Vuuren's is a technocrat's line, though, because in fact most people's tools for judging water quality aren't up to the task. Conventional wastewater treatment is very good at removing the kind of contaminants people can detect without laboratory equipment, such as odors, suspended particles, and the sorts of bacteria that can cause illness. But most people are relatively helpless when it comes to making more detailed assessments of their water supply's safety. The lower Colorado River looks clean enough; it's more likely to meet most people's standards than cleaner water in a pipe outside a complex-looking treatment plant.

As a result of that perceptual shortfall, people are left with nothing but water's history as a guideline, according to Brent Haddad of the University of California at Santa Cruz, an environmental studies professor who directs the university's new Center for Integrated Water Research. When he began studying water policy, he says, "I kept going to meetings with water managers, and they kept saying, 'How do we deal with these irrational people?' - meaning their customers. I didn't think they were irrational. I thought they were just using a different sort of logic than the water managers and engineers. People as they are generate feelings and opinions about some things that are really based on intuition and not a technical analysis of risk. They're based on what you might call ancient rules of thumb about what's safe and what isn't."

A visceral aversion to unclean water, Haddad says, is an understandable and useful tool that served the human species well through most of its evolution. But it may not be particularly helpful today, when it's necessary to make a decision between two sources of water that are both clear and odorless - but from very different sources.

 "When people are aware of the history of their water, it matters a lot to them," he says. "If there's an unavoidable link to prior urban use, that's troubling to people. It's extremely hard to convince people then that the treatment will be good enough to override that history. But people are willing to take Colorado River water or groundwater that's clearly been used by other cities because it's easy to abstract away that use and begin the water's history with its taking from the natural system."

Rivers and soils do, in fact, clean water. But the psychological cleansing they do may be equally important. As a result, even the Colorado River - however thoroughly dammed, diverted, and delivered through aqueducts it may be - appears more natural, and cleaner, to many people than what's produced by San Diego's wastewater treatment plants. The river takes the yuck out.

The largely unwelcome prospect of drinking treated effluent, though, forces people to ask what's in the water they're already getting, whatever its source. Something long taken for granted - what could be more American than good, drinkable tap water? - becomes a public issue. And as people debate where their future water supplies are going to come from, an increasing number of experts and nonexperts alike are growing increasingly alarmed about the chemicals flowing not only from Las Vegas, but from every community.

 Wastewater engineers are rightly proud of what their industry achieved in the 20th century, bringing safe drinking water to virtually every community in the United States. But most wastewater treatment plants were not designed to remove the sorts of complex organic chemicals that show up in Lake Mead - or, to cite a more pristine-looking example, Boulder Creek, which tumbles out of the Rocky Mountains and through Boulder, Colo., before joining the South Platte River.

Back in 2000, David Norris thought Boulder Creek an unlikely place to look for unhealthy fish. Even below the city's wastewater treatment plant, the creek looked clean,

and fish and other aquatic organisms lived throughout it. There was none of the stench, the brown murk, or the belly-up fish associated with the bad old days of piecemeal sewage treatment before the Clean Water Act was passed in 1972.

Norris, an endocrinologist at the University of Colorado - and an avid fisherman - had read studies in the scientific literature documenting the environmental effects of a poorly understood class of pollutants known as endocrine disruptors. Unlike many toxins, they didn't appear to be killing their victims outright. But in Lake Apopka, Fla., a pesticide spill had caused lingering reproductive failures and sexual abnormalities in alligators. In Britain, odd-looking fish that were not readily identifiable as males or females, but had sexual characteristics of both, were turning up in anglers' creels - especially in waterways below sewage outlets.

Norris and his colleagues, Alan Vajda and John Woodling, figured that Boulder Creek's best indicators of environmental quality were likely to be white suckers, a native fish that's widespread and not terribly finicky about water quality. "A good healthy freshwater stream has a good healthy sucker population," he says. "If you really disturb this species, you've really disturbed the ecosystem."

Norris had no trouble finding white suckers both upstream and downstream of Boulder's treatment plant. Upstream, everything seemed normal. Downstream, it was not. "Much to our surprise," he says, "we were appalled to see the extent of feminization in the fish population." He found five female suckers for every male; further, 20 percent of the fish were "intersex" individuals showing characteristics of both sexes.

Alarmed, Norris looked for similar effects elsewhere, and found them. Fish below wastewater treatment plants in Denver and Colorado Springs showed some of the same symptoms. In the South Platte River, where Denver releases its waste, he couldn't find a single male sucker below the effluent outlet. Something in the effluent, it appeared, wasn't killing fish, but rather causing hormonal changes in them and producing female traits in male fish.

The evidence was circumstantial, though. Norris knew he had to more closely link cause and effect - which is hard to do in a natural setting, where fish in different reaches of the same stream might be feeding on different food, facing different temperatures, and otherwise dealing with widely variable conditions. So he and his colleagues have since built two "Fish Exposure Mobiles," which are basically mobile laboratories, built inside trailers, with fish-holding tanks. By pumping combinations of river water and wastewater effluent into the tanks on site, they're able to replicate the pollution concentrations fish face at various distances below treatment plants.

When they experimentally exposed fathead minnows - widely used as a test fish - to water like that below the Boulder treatment plant, Norris and his colleagues were able to feminize male fish within 14 days. They have since tested fish in other Colorado waterways below wastewater treatment plants in the Rocky Mountains and on the Western Slope. Data from those tests aren't available yet, but Norris will say that he is awfully worried in general about the presence of endocrine-disrupting chemicals in the environment, and in water specifically.

 "It's fairly obvious that living populations are being subjected to far more chemicals in the last 30 years than when biological systems evolved, and so we wonder what effect that has on the genetic machinery," he says. "If we want to increase the use of wastewater, unless we're going to remove these compounds from the water, we're going to increase their concentration in the human population, since we're just going to be adding more of these compounds. We keep concentrating our population in cities, and as a result we're concentrating our effluent."

Most of the organic compounds that can disrupt the endocrine system are neither regulated by EPA standards nor often monitored in waterways or the drinking-water system. Few thought they were a problem until recently. But in a national survey published by the U.S. Geological Survey in 2002, researchers found such substances in 80 percent of the waterways they sampled.

The endocrine system is essentially a complex signaling mechanism that tells genes and cells when to do what. It operates by means of chemical messengers, or hormones, that bind to certain receptors in cells. Unfortunately, many of those receptors aren't particularly picky. Receptors designed to react to the natural hormone estrogen, for example, can also be set off by a wide range of other compounds, from complex molecules that naturally occur in vegetables to synthetic chemicals found in soaps, plastics, pesticides, cleaning products and many of the other manufactured goods of modern civilization. They get into sewage when people urinate, or shower, or flush leftover pharmaceuticals down the toilet.

As in Boulder Creek, waterborne endocrine disruptors have in many places been shown to have harmful effects on aquatic organisms, especially fish. For example, male carp with unusually high levels of female hormones have been found in Lake Mead, where estrogen - the kind naturally produced in human bodies as well as the synthetic variety in birth-control pills - ends up when Las Vegans flush their toilets. Recently, a team of Canadian biologists dosed an entire small lake with synthetic estrogen at levels equivalent to those often found in treated wastewater. They were able to wipe out almost the entire minnow population in only a few years - again, not by killing the fish, but by causing sexual changes in males and females that made it impossible for those fish to reproduce.

Hormones naturally work at very low levels; a human estrogen concentration as low as 1 part per trillion - so dilute that it's near the lower limit of what monitoring equipment can detect - has been shown to affect fish. The effluent dumped into Boulder Creek typically contains from 1 to 10 parts per trillion of human estrogen.

"People ask why such tiny levels have such a devastating effect," says Norris. "But that's the level at which hormones work. Parts per trillion is common stuff for an endocrinologist."

Consumers are used to thinking of drugs as having precisely tailored effects. But endocrine disruptors don't work that way. Because many different chemicals can activate a given set of hormone receptors, low doses of quite different substances can combine into a higher dose. That's one of the primary reasons a growing number of researchers worry about possible implications for human health. 

"What happens when you have a summing-up of the effects of these different chemicals?" asks Theo Colborn, a longtime pollution researcher who runs the nonprofit Endocrine Disruption Exchange in Colorado and coauthored the  1996 book Our Stolen Future, one of the first popular publications to raise an alarm about such compounds. "Some-times there's even a synergistic effect between them. It's like adding 2 and 2 and getting 5."

 Wastewater treatment lowers concentrations of most trace organic compounds - often by an order of magnitude or more - but it can't remove them all. As a result, effluent often contains a stew of complex chemicals. A recent U.S. Geological Survey study found that St. Vrain Creek, into which Boulder Creek drains, carries measurable loads of at least 36 different compounds, including artificial fragrances, fire retardants, antibacterial substances used in soaps, and substances used to manufacture plastics. The extent to which those chemicals work together to cause effects on the endocrine system - itself not well understood - is a big unknown.

 "The endocrine system is much more than estrogens," says Catherine Propper, an endocrinologist at Northern Arizona University who has studied the effects of trace organics on amphibians. "We have this complicated endocrine system, and every time we find new aspects of it, we find they can be disrupted by some of these environmental contaminants."