(FORTUNE Magazine) – In 1956 a grouchy, iconoclastic shell Oil geologist named M. King Hubbert forecast that U.S. oil production would peak by the early 1970s. The general reaction to his prediction was disbelief and ridicule--until, that is, the early 1970s, when he turned out to be right. Disciples of Hubbert, who died in 1989, have continued his work and are producing more bell-shaped curves that resemble his initial chart, which is known as "Hubbert's peak." Only now Hubbert's followers are plotting worldwide production, and in doing so have come to an uncomfortable yet oh-so-logical conclusion: We are running out of oil sooner than conventional wisdom would have us believe. Indeed, peak worldwide oil production may be at hand.

Before you dismiss the Hubbert crowd as fringe-element doomsayers, remember that the same label was attached to Hubbert himself for decades. Hubbert's science is real enough, and of course it is unassailable that at some point we will run out of petroleum. What's disputed are estimates of supply, consumption, and undiscovered oil. Some call Hubbert's analysis simplistic. That's fine by Princeton professor emeritus Kenneth Deffeyes, who once worked with Hubbert at Shell and is the author of Hubbert's Peak. "Use the simplest hypothesis that fits the data," Deffeyes says. "Hubbert's hypothesis fits the world data quite well, thank you. 'Simplistic' is a virtue."

Geologists long believed that oil consumption and oil discovery would continue to grow at the same rate forever. Hubbert was the first to make the obvious point that that would eventually prove false. In his new book, Out of Gas, Caltech professor David Goodstein shows that Hubbert's calculations explain how once the rate of discovery slows, it's possible to extrapolate where production growth ends. And the rate of discovery worldwide has been in decline for decades. Bottom line: The folks in the Hubbert camp generally believe that the total amount of oil that was ever on planet earth is two trillion barrels and that we have gone through nearly half of that--which puts us close to Hubbert's peak.

Yes, it's true that tabulating world oil reserves is a fuzzy business. OPEC countries, for instance, tend to overcount so that they are allotted higher production quotas. And oil companies fudge to please investors (as the aforementioned Shell Oil Co. has). Also, technology improvements make more oil economically accessible, which adds to the count. All that, Goodstein points out, may delay Hubbert's peak. But what's important to note is that the Hubbertians believe the crisis comes not when the last barrel of oil is pumped but when the peak is attained. That's when the price shock hits (as in the '70s).

One of the scariest points I've seen made in response to Hubbertianism is the line that there's plenty of oil out there because little exploration has been done in Iran, Iraq, Kuwait, Saudi Arabia, and the United Arab Emirates over the past 25 years. Great! In other words, those countries will hold even more sway over our future energy needs!

Oil may not stay at $40 a barrel forever. But Deffeyes and Goodstein, among others, argue that America needs a new, forward-looking energy plan to furiously explore and develop other energy sources and models. Hubbert himself talked along those lines in a 1978 AP interview. As usual, he was a few decades ahead of his time: "The federal government has to go all-out on a program of solar energy--on the magnitude of what we did in developing an atomic bomb or the space program," Hubbert said. Is weaning our economy off oil--particularly imported oil--as important as winning World War II or sending a man to the moon? You don't need to be an iconoclastic grouch to believe that today.

ANDY SERWER, editor at large of FORTUNE, can be reached at aserwer@fortunemail.com. Read him online in Street Life on fortune.com and watch him on CNN's American Morning and In the Money.