I’ve been mindful of the amount of water I use when making a pot of coffee ever since learning that one-third of the tap water used for drinking in North America is actually used to brew our daily cups of joe—and that if each of us avoided wasting just one cupful of coffee a day, we could save enough water over the course of a year to provide two gallons to every one of the more than 1.1 billion people who don’t have access to freshwater at all.

That is a stark statistic, when as many as 5 million people die unnecessarily each year because of lack of water and water-related illnesses; one-third are under age 5.

So for me that excess cold coffee at the bottom of the pot became a bothersome reminder. But I had never thought beyond that—about how much water it takes to actually grow the coffee. That amount is called virtual water (pdf), and it’s the kind of thing you don’t really consider until someone brings it to your attention: “Do you know how much water it took to make this?” Virtual water is a calculation of the water needed for the production of any product from start to finish.

Here’s how it is figured: It takes about 155 gallons of water on average to grow a pound of wheat. So the virtual water of this pound of wheat is 155 gallons. For a pound of meat, the virtual water is 5 to 10 times higher. There’s a virtual water count for everything. The virtual water footprint of a cup of coffee is 37 gallons; an apple, 19 gallons; a banana, 27; a slice of bread, 10; a sheet of paper, 3; and a pair of leather shoes, 4,400, according to Waterfootprint.org, a Unesco-run Web site providing a calculator for individual and national water use. In fact, virtual water in internationally traded food and products such as these accounts for 15 percent of global water consumption.

Virtual water matters a lot these days because we are in an encroaching global water crisis. According to the United Nations Environment Programme, the world needs to increase its water supply (pdf) for irrigation by 14 to 17 percent by 2030 just to meet its dietary needs. Virtual water is where major savings can accrue.

Proper management and use of the world’s virtual water already save almost 5 percent of the water used annually in global agricultural production, according to Unesco. This follows a simple logic: Places with less water gain access to foods with high water requirements by importing them from areas with high rainfall or substantial water supplies. This allows water-scarce regions to use their own water resources more efficiently for other purposes—and create water savings. For instance, areas of southern China that have more water and are better equipped to grow certain water-intensive agricultural products can send them to northern China. This frees up northern water supplies for other uses, such as drinking and sanitation. Jordan saves 60 to 90 percent of its domestic water supply by importing water-intensive products.

The water savings are even greater than they seem at first. Producing grain and other foods in an arid country like Jordan may require two or three times the water it takes in humid settings in South America or the United States. So the virtual water saved may be three times the amount that was actually necessary to grow the crop in a more appropriate climate. It’s all about being smart with water.

Yet we can be smarter, and need to be.

Right now we lose 30 to 50 percent of the food we grow—and all the virtual water in it—by the time it is ready for consumption, says Daniel Zimmer, executive director of the World Water Council (WWC) in Marseille, France. These losses come in harvesting, production, processing, transportation, and storage. Tossing out leftovers wastes every drop of water it took to grow the food (and think of all the times you don’t ask for a doggie bag). Indeed, the third most common refuse found in landfills is food, according to the Environmental Protection Agency. “Sure, a few liters of water are saved when you take a shorter shower,” Zimmer says. “But hundreds of liters of water are lost when you throw away food. We have to begin to think about our water use differently.”

I like the idea of virtual water because it helps us think about our water use differently without having to make giant, complicated leaps. It puts water into the context it deserves: We use freshwater mostly for agriculture, not for drinking or bathing. Today agriculture accounts for about 70 percent of all water use in the world and up to 95 percent in several developing countries. So it makes sense to first start looking at savings via food production. And when I say savings, I mean efficiencies and better water management, not necessarily avoiding particular food groups altogether—although that isn’t such a bad idea once in a while either. Meat requires 5 to 10 times more water to produce than vegetables do. Swap the two in your diet and you will save up to 750 gallons of water a day. (See “What’s Your Virtual Water IQ?” page 26.)

While thinking about water differently should be a moral imperative, in a world view it comes with controversy. “At the global level, virtual water and the trading of it has geopolitical implications,” the WWC says in a report on the subject (pdf). “It induces dependencies between countries....This can be regarded either as a stimulant for cooperation or as a reason for potential conflict.”

Right now the United States is a water exporter, but population growth, pollution, and lingering drought in vast regions may change that. “As demand grows we are going to have to ask what is it being used for and whether that is a good use of our water,” says Maude Barlow, cofounder of the Blue Planet Project. “One-third of the water in the United States is exported as virtual water when a number of major water systems in the United States are in a catastrophic decline. People may begin to say, ‘Why are we shipping our water away?’”