(FORTUNE Magazine) – MCI WorldCom CEO Bernie Ebbers is sitting at his desk in Clinton, Miss., talking with a visitor about his favorite subject--okay, everybody's favorite subject these days, the stock market--when he swivels around to peer at a laptop flashing the day's share price. He squints and finds what he's looking for. "We've got this idiot giving a press conference on the WorldCom/Sprint transaction," Ebbers groans. The "idiot," a regulatory analyst, believes the Department of Justice will block Ebbers' $129 billion deal to buy Sprint. As Ebbers expected, WorldCom has dropped two points since the conference call began.

WorldCom doesn't quite cut the same heroic figure it did a couple of years ago, at least not on Wall Street. But if Ebbers is worried, he doesn't show it. He believes he is on the cusp of closing the biggest merger in U.S. corporate history. If he pulls it off, he thinks he can end the Baby Bells' lock on local phone service, rule the cellular airwaves across the U.S., and show the Europeans how to sell phone service in Europe. Oh, and he's going to invest $100 billion in the Internet.

From an improbable origin as owner of a small chain of motels, Ebbers, 58, has become the person who sets the tone for the global telecom industry. First using profits and then using stock, he has bought companies by the dozen, always paring costs to the bone to keep his margins high and his shareholders happy. He rolled up operators like MFS and Brooks Fiber, which owns networks in big cities, and UUNet, which has the largest Internet backbone. As telecom consultant Mark Bruneau of Renaissance Strategy Worldwide puts it, "WorldCom is the poster child for buying real things with fake money."

By 1998 his scrip was worth enough to pay $40 billion for MCI, a company three times WorldCom's size, increasing revenues to $37 billion last year. That, along with his plans to build a new high-speed network in Europe, gave the company a focus: It would court business customers as the specialist in carrying phone calls and data anywhere in North America and Europe--almost all on its own wires. WorldCom would be the first truly global phone company.

And Ebbers would be the man most adept at adapting to and even defining the swift changes in technology and regulation that have energized the industry. Global Crossing CEO Leo Hindery says, "Ebbers is a rock & roll star. Now telecom is like the entertainment business, where you know the CEOs just by their first names."

Soon after MCI joined the fold, though, came the telecom industry's equivalent of a stretch in rehab. MCI's dial-around and consumer long-distance businesses sagged as price-conscious customers churned in and out and began to make long-distance calls on their wireless phones, a business in which neither MCI nor WorldCom has ever had much presence. Even worse, WorldCom's wholesale revenues from selling service to other carriers fell by 16%, forcing Ebbers to cut his forecast for growth in total revenues by a couple of points, to 15% or 16%--still quite good but not as good as before. "When you do that, stockholders don't find that very encouraging, and rightly so," he says. "You had better swallow deep and get your act together and get on with it."

So Ebbers did what he has done before: He went out and made a deal to change the company again.

Sprint will help WorldCom in ways both obvious and unexpected. Global corporate customers are the sweet spot in the telecom racket today, and WorldCom faces a stiff contest. AT&T and BT have a joint venture called Concert expressly to chase these customers, and when Leo Hindery, the Global Crossing CEO, was asked to name his Holy Grail, he responded without a moment's hesitation: "MNCs"--multinational corporations. The market is wide open too. As Sprint CEO William Esrey says, "Nobody is doing a great job of serving these customers." He should know. Sprint was part of Global One, an ill-starred alliance with France Telecom and Deutsche Telekom that disintegrated last year.

Ebbers has one big advantage--his own international high-speed network. But as he himself puts it, "The question is, How much of an advantage is it?" Qwest and Global Crossing are building overseas too. If they can lease space to other carriers and make it easy for them to turn on service for business customers, WorldCom's advantage can fade. As consultant Mark Bruneau says, "Bizarrely, it's not a network game. It's a marketing game."

That's where Sprint comes in. According to telecom analyst Lawrence Kenny of PricewaterhouseCoopers, Sprint is stronger than WorldCom in marketing to corporate customers and has a big presence in Latin America. Kenny says, "If you want to focus on the FORTUNE 2,000, scale is important."

But even as it helps WorldCom in the bread-and-butter corporate market, the Sprint deal has more to do with something else. The twist is that it's really about letting WorldCom attack a market that isn't fashionable among telecom executives: consumers. Ebbers is betting that changes in technology and lifestyle will hobble a telco that ignores the residential market.

Ebbers needs consumers because he is facing capital demands of almost laughable proportions. He wants to be king of the Internet, and winning that job will be expensive. At the rate Internet traffic is growing, he argues, the cost of beefing up the network with new electronics and software over the next four years will be $100 billion. For the industry as a whole? "No," he says. "Just for WorldCom." Ebbers insists the figure is real, though WorldCom's capital budget this year is merely $8 billion. But even if new efficiencies in technology let him add the capacity he needs for a lot less than $100 billion, he's talking serious money.

Consumer profits would help fund this investment. Phone companies face the same economic conundrum as any enterprise that makes its money by moving people or things from one place to another: They must invest enough to carry all their traffic at hours of peak demand, even though that means laying on capacity that will lie dormant the rest of the time. If the peak is tall and narrow--as for a telco that carries calls from businesses open from 9 to 5--the transport company must hustle up traffic at other times to help pay for overhead. For WorldCom and its competitors, that means either selling space in bulk to other carriers--a business that tanked last year as new long-distance companies flooded the market--or finding and keeping residential customers, who make most of their calls after business hours.

That has proved difficult for WorldCom. Because it has no cellular systems, when MCI customers use cell phones to make calls while away from home, the money they spend comes straight out of WorldCom's hide. This hole in WorldCom's portfolio will grow bigger as cell phones and other wireless devices begin to replace the PC as the main point of entry to the Internet. At the same time, the ubiquity of wireless gadgets is blurring the line between work and leisure as individuals use the same devices in both home and office. If WorldCom is to hang on to its mobile long-distance customers and be a player when wireless data traffic explodes, Ebbers needs to plug a leak, and fast.

Buying Sprint would do so in a stroke, because it has a nationwide cellular network, Sprint PCS. Unlike other big networks, which are a pastiche of analog and digital systems, Sprint's network is standardized on a single digital technology and operates over the same frequency nationwide. That makes it well suited for carrying Internet and other data services.

But an equally intriguing facet of the Sprint deal, and possibly a more far-reaching one, involves a little-known wireless technology called MMDS, which may be Ebbers' secret weapon in his assault on the consumer market. MMDS, or multipoint multichannel distribution system, was conceived as a way to broadcast lots of video channels in a single burst. But the spectrum the FCC set aside for MMDS has for the most part lain fallow. Many licenses went to nonprofits like the Catholic Church, which didn't have much of a programming budget, or to independent companies hoping to compete with cable-TV operators, only to find that they couldn't squeeze enough channels into their allotment.

Eventually the licenses fell into the hands of two rather large telcos: MCI and Sprint. A few years ago, Sprint quietly started a research project code-named Mighty Mouse to see whether the MMDS spectrum could be put to a different use: two-way communications for phone calls and high-speed Internet traffic.

Sprint found that it could beam its signals from a central broadcast tower to a small flat antenna mounted on the side of a house. A coaxial cable would carry the signal to a decoder box in the basement, which would feed it to the right telephone or computer upstairs. Anyone within 35 miles of the central mast could use the system as long as the receiving antenna was within direct line of sight. Sprint began buying its licenses about a year ago. So did WorldCom, with the same end in mind.

If the companies can complete their merger, Ebbers will control MMDS spectrum that can reach about 60 million households. Rolling out MMDS would put him on the same footing as local phone companies and cable-TV operators, forging a third link into the home, albeit a wireless one. As Bear Stearns telecom analyst William Deatherage says, "The combined company could break up the duopoly in the local market. That's a real fork in the road."

Provided that the strategy works. Wooing consumers with a new local service may distract Ebbers from his courtship of business customers, and technical standards for MMDS systems have yet to be finalized. Ebbers also faces a big decision about what services to offer over MMDS. Sprint is pushing a high-speed digital service called ION for integrated on-demand network. ION gives customers a fat digital pipe that they can fill with phone or data services in whatever combination they see fit. But ION will be expensive--$159 a month for unlimited local calls, unlimited high-speed Internet access, five e-mail addresses, and 750 minutes of long distance. AT&T, on the other hand, just began selling local phone service and high-speed Internet access for only $60 a month over a new wireless system of its own in Fort Worth. If Ebbers hopes to make a big dent in the local market, he may need to step back and go down-market.

Of more immediate concern to Ebbers is getting the deal done. The Department of Justice and the Federal Communications Commission each must rule that the deal does not hinder competition. By far the greatest sticking point is the strong position of WorldCom and Sprint in the Internet backbone business--Nos. 1 and 2, respectively, in traffic handled. So it is taken as a given that Sprint will have to sell its Internet backbone business--including customers, employees, and the electronic hardware that feeds traffic onto the Net. MCI had to do the same to win approval of its merger with WorldCom two years ago.

And there's the rub. The sale of the MCI backbone to Britain's Cable & Wireless was by all accounts a disaster. According to regulatory analyst Scott Cleland of the Legg Mason Precursor Group--the one Ebbers complained about--Cable & Wireless did not get a full list of its customers when it took over the business and still has not received all of the contracts. MCI transferred only 43 salespeople to manage 5,000 accounts. Eventually C&W sued and settled for $200 million in compensation.

Cleland says the divestiture went poorly partly because MCI was reluctant and hence unhelpful. The bigger problem is that the Internet backbone was tightly intertwined with the long-distance backbone, sharing customer base, infrastructure, back-office systems, and sales force. Cleland says, "You couldn't say whether a salesman or an engineer was an Internet-backbone person or a long-distance-backbone person."

Sprint will face the same sort of challenge when it tries to hive off its Net business. To date, the company has not even broken out those revenues. Nor does it help that one of the most vociferous critics of the MCI/WorldCom deal was Sprint, which complained that the merged company would control more than half of all Internet connections and would "very likely short-circuit the growth of the global information network." Sprint CEO Esrey says with characteristic wryness, "People understand why we took that position. But it doesn't go in the plus column."

Even so, most analysts think the deal will win approval, arguing that the botched MCI divestiture will help regulators know what to look out for. Sprint President Ron LeMay says Sprint will be able to persuade regulators it can deliver "by demonstrating a level of detailed planning that evinces an understanding of the issues." To that end he is hastening to put the Internet backbone into a separate, newly created business unit. The Justice Department is expected to issue a ruling by July.

If Ebbers, Esrey, and LeMay can navigate the bureaucratic shoals, next will come a period of cultural bonding. For Sprint, that may end up feeling more like bondage.

Clinton, Miss. (pop. 21,847), might seem like a quiet town for a global telecom company, but it holds an interesting place in American history. A few miles from WorldCom's showy new steel-and-glass headquarters is the site of a World War II camp for German prisoners of war. Its interns provided the nonvolunteer labor force for the first phase of the Army Corps of Engineers' Mississippi River flood-containment project. Camp Clinton was also the place where nearly every German general captured by the Yanks finished out the war.

WorldCom is in Clinton because Ebbers likes it there. Mississippi College, his alma mater, is just across the road; the college sold Ebbers the land for its new edifice. But only a few hundred people work in the Clinton headquarters. When Ebbers buys companies, he leaves his new employees where they are. Sprint will stay on the new campus that Esrey is building near Kansas City.

All the same, life will change for the folks at Sprint. If the past is a guide, they had better be prepared for the laser light Ebbers will shine on their costs. After the last merger, Ebbers gave MCI staff a bit of a shock when he made them stay at cheaper hotels and got rid of company cars. Suppliers get the Ebbers squeeze too. One consultant says that as soon as the deal goes through, Ebbers will seek to renegotiate contracts. "They're a devil to get money out of," says the consultant, "and they price-shop every single input, all the way through to envelopes and paper clips and, unhappily, management consultants." Esrey says, "People won't stay if it's an environment they're not content to work in." But he immediately adds, "All of our relationships have been terrific."

In Esrey and LeMay, Ebbers will join up with two of the industry's more interesting characters. Esrey, 60, an old AT&T veteran who looks like a slim Daddy Warbucks, is candid and self-deprecatingly funny, traits he shares with Ebbers. He has been CEO of Sprint since 1985--a long run in this business, especially since not all his bets have worked out, as in the Global One joint venture and Sprint's investment in Iridium, the bankrupt satellite-telephone consortium. But he's had some huge successes too, as in late 1998, when he divided Sprint's shares into two tracking stocks for its wireless and regular phone businesses. While its regular stock has limped along, shares in Sprint PCS are up 700%.

Esrey has yet to settle on a post-merger role. He will be called chairman, but chairmen have a way of disappearing at WorldCom, and Esrey doesn't want to be invisible. His predecessor, former MCI CEO Bert Roberts, was in charge of aircraft and the foundation and not much else. Ebbers will need to find more for Esrey to do if he wants his new colleague to stick around.

His plans for LeMay, 54, are clearer. Ebbers has always been WorldCom's de facto chief operating officer, with all units reporting directly to him. "I don't want to do as much as I have," says Ebbers, "and I keep my fingers crossed that Ron is going to want to do a lot of it." LeMay will be called chief operations officer--that's "operations," not "operating." Ebbers reportedly feels the latter title would designate LeMay as his successor, and he doesn't know him well enough yet.

LeMay appears at first glance to be rather strait-laced--not the sort of guy you'd think would let himself be strapped to the inside of a big, hollow, metal ball and rolled around the floor at a sales conference. But he did, at Sprint, six or seven years ago. In conversation he can be intense and focused, but then he'll break out in an exuberant cackle. "I am responsible for periodic pranks," he admits, "and Esrey is the most satisfying person to play them on"--a caveat emptor if ever there was one.

At least Ebbers seems like someone who can take a joke. He is certainly good at giving them. He excels at the art of delivering sharp teases that somehow land without stinging too hard, a favored target being people he towers over. Addressing the dinner audience at a recent New York City telecom conference hosted by First Boston analyst Dan Reingold--who is about a head shorter than the WorldCom CEO--Ebbers said, "They put these podiums low enough for Dan to see over. But I can't see my papers."

What's most notable about Ebbers' demeanor is how unfazed he seems by it all. Since FORTUNE visited him in Mississippi two years ago, his company has quintupled in size and his stock has gone to hell, but he seems just as relaxed and at ease, if not more so. His readiness to bring up WorldCom's crummy share price without any prompting is typical of his confidence. He says that he's not thinking about retirement, but that the board of directors will probably get tired of him after a while. Then he jokes, "With the stock down the way it is, it may happen sooner than we think." Now that he's a telecom mogul--a label that makes him cringe--he has moved into a real house across the street from the double-wide mobile home he used to inhabit. But he still blows off steam by puttering around his farm on a tractor.

For fun, he likes to dabble in the market, and on the day that Scott Cleland's conference call is sending WorldCom's stock down a couple of points, he checks another ticker. "Inktomi is up $17," he says, shaking his head in disgust. "I'm so stupid!" Ebbers had bought a few shares a little while before and sold them at a modest profit. Then he saw the price swinging up and down and thought that he could play with the stock between $100 and $120. He bought some more and told his broker to sell when it rose to the top of the range. "I got out at $119," says Ebbers. "Then it started going down again, and when it reached $108, I told my broker to buy when it fell to $105." He pauses. "It never got there. Now it's at $217. And this in four weeks!"

It's a cruel game. But who can always know the right time to take a chance?