(FORTUNE Magazine) – In early 2000, a company called Plug Power charged out of obscurity (okay, upstate New York) into stock market stardom. Plug is developing something called a fuel cell, and there's talk that within a few years, fuel cells, which are actually more than a century old, will revolutionize the way we power our cars, our homes, our entire lives. Since there's no way to quantify Plug's true potential--actual products, much less profits, are still somewhere off in the future--Wall Street has decided that it must be infinite. Plug's stock exploded from its $15 October IPO price to a high of $150 in mid-March; the stocks of other fuel-cell developers, like Ballard Power, surged as well.

As the rise and fall of one dot-com after another makes clear, you can't trust this market to distinguish fantasy from reality. Today fuel cell stocks have plummeted back to earth. So where does the truth lie?

Fantasizing about fuel cells has been popular sport since 1839, when British scientist Sir William Grove discovered that if you combine hydrogen and oxygen electrochemically to form water, you get heat and electricity. International Fuel Cells, a division of United Technologies, has been working on fuel cells since the 1950s; its products have powered everything from Apollo 11 to a police station in the middle of New York City's Central Park.

But IFC's current fuel cells (which are really stacks of cells, since a single cell produces less energy than a flashlight battery) are large (18 by ten by ten feet) and expensive. The real excitement--and the current focus for companies like Plug Power, Ballard, and IFC--lies in bringing fuel cells to the masses. Big reductions in cost over the past decade are for the first time making that a possibility. Combine a few cells, and you could have energy for, say, a laptop. Stack a bunch together, and you could power a house, or a car. If that proves economical--a big if--then fuel cells could be a $100 billion market by 2010, says analyst Bobby Winters at Bear Stearns.

That potential is why some serious money is backing fuel-cell research. General Electric owns 12% of Plug. Ballard Power is 32% owned by DaimlerChrysler and Ford Motor. Indeed, Winters estimates that automakers have spent upward of $1 billion on fuel-cell technology. In January, Bill Gates set off a rocket when a regulatory filing showed that he had taken a 5% stake in Avista, a 113-year-old Washington utility whose subsidiary, Avista Labs, is developing fuel cells. Avista's stock soared 46% in one day.

The excitement is well founded. Fuel cells fit today's focus on clean energy, since the frictionless, efficient process throws off few nasty byproducts. IFC claims that using a fuel-cell system to power an office building can save more than 40,000 pounds of air pollution a year. And fuel cells could be a huge boon to people living at the outer edges of the so-called grid. Plug Power CEO Gary Mittleman says that about 18 months ago a farmer drove up to Plug's Latham, N.Y., headquarters in his pickup. He said to the receptionist, "Honey, I'm here for my fuel cell. Where's the loading dock?" Then he dropped a brown paper bag with $10,000 in cash on her desk. That seemed a small price to pay: His livelihood is threatened when the power goes out on his remote Vermont farm.

But since Plug is about a year from a commercial product, the farmer went home disappointed. All the fuel-cell companies claim to be very close to delivering commercially viable cells. Ballard says that portable products--like Honda generators powered by Ballard cells--will be on the market in 2001.

But even with their stock price corrections, these companies still must conquer the residential and automotive markets to grow into their valuations. People expect their home energy supply to be cheap and reliable. Powering cars is even more complicated. How long will it be before you'll be buying a fuel-cell-powered car at your local Web dealership? At least a decade, say most observers.

That's why the current hype could backfire badly. Witness Plug. In early May it disclosed that GE, which was contractually bound to buy some 500 prototype cells this year, is not so bound, because of design and timing changes. The companies say their relationship is intact, but Plug's stock plunged 27% in one day. "We want to encourage our colleagues to deliver on their promises. Don't set expectations too high," says Avista Labs President Kim Zentz. It may be too late for that.