(FORTUNE Magazine) – It seems the perfect antidote to the often painful process of buying a car. "Welcome to John Elway AutoNation," says the smiling man at the door of the Ford dealership in Denver. "How may we help you today?" In a flash, you are in the custody of a polo-shirt-and-khakis-clad "sales consultant," who is friendly and persistent but not obnoxious. Not once are you asked the car peddler's loaded question, "What would it take to make you buy today?" That's because the price of every car on the lot, new and used, is fixed. There's no haggling, no "let me take your offer to my manager" gamesmanship--but also no discounts. The sales consultant couldn't change the price if he wanted to. He can't even throw in free floor mats.

When it comes to selling cars, the 17 John Elway AutoNation USA dealers in the Denver area represent the state of the art--and Wayne Huizenga's best, perhaps final, shot at cracking the code of auto retailing. The man who made himself a billionaire at Waste Management and Blockbuster has been selling new and used cars for three years now. And although he has changed his strategy nearly as often as Kate Moss changes clothes, he hasn't yet figured out how to make money at it. "There are a lot of moving parts in this business," Huizenga says. "It isn't a simple thing like renting a video or picking up the trash. It is difficult, but you have to keep working at it."

Huizenga's not the only smart businessman to crumple a fender against the decades-old auto-retailing system. New York financier Marshall Cogan expected his United Auto Group to create vast economies of scale; in fact, all it created were substandard profits and bitterly disappointed shareholders. Austin Ligon, formerly a management consultant with the Boston Consulting Group, has yet to turn an annual profit with CarMax. Even the Internet, which was supposed to change everything, has scarcely made a dent in traditional automotive retailing practices.

It's not for lack of trying. Besides instituting no-haggle pricing, the retail reformers have cleaned up their showrooms and added coffee bars, nurseries, big-screen TVs, and putting greens. They have installed computer-driven kiosks where customers can get information about the car they want without talking to a salesperson. They offer money-back guarantees: three days or 150 miles. But it has all been to little avail. The much-derided plaid-jacketed fast-talking car salesman, the one who antagonizes customers but moves a lot of metal anyway, remains at the center of most successful retail operations.

Why this should be so is one of the mysteries of American retailing. If any sector cries out for a Sam Walton or a Jeff Bezos to come in and overturn the old ways, this is the one. It is no joke that many Americans would rather go to the dentist than shop for a car. They drive off the lot not knowing whether they got the deal of a lifetime or were taken to the cleaners. "The most successful dealers are the ones that are not trying to change the consumer proposition at all," says one veteran auto executive, who asked not to be named for obvious reasons. "They are entrepreneurs, and they go right to the edge: 'How little can I give you on your trade-in, and how much can I deceive you about your monthly payment?' It becomes anything but consumer-friendly."

Improving on this arrangement wouldn't appear to be beyond the abilities of the corporate minds that brought you Wal-Mart, say, or Amazon.com. That's especially true these days, when the auto industry is enjoying its greatest period of prosperity ever. This year Americans are buying new cars and trucks at the fastest pace in history, close to 17 million annually, and demand far outstrips supply for many hot-selling trucks, sport utilities, and luxury imports. It would seem to be an ideal time to impose some rationality on the system and make money at the same time. Certainly the traditional operators, protected by rigorous state franchise laws that limit competition, have had no problem getting richer. The talk around the industry is that if you want to find a successful local dealer, don't look at his place of business. Go find him on his boat or at the country club.

But there's clearly more to running a profitable dealership than the men in suits ever bargained for. In addition to selling new and used cars, you also need to manage a parts business and a service operation, two quite different propositions. The trickiest part is inventory control: making sure the right models are in stock and then adjusting prices to reflect the market, competition, and the cars' physical condition. (Huizenga, for instance, reprices new cars every 20 days and used cars every 15 days.) Corporate America's continued inability to make a go of all this reinforces many car people's prejudice that successful dealers are born, not made.

If anyone is determined to make himself into a successful dealer, however, it's Wayne Huizenga. He's already the biggest car seller in the country, with 287 outlets that are expected to produce $20 billion in revenues this year. And he has already tried--and rejected--a wide range of new strategies. He started with the category-killer model, building multi-acre used-car superstores around a metropolitan region, all served by a central used-car reconditioning center. That didn't work because the huge car lots cost too much to build and because the sales force couldn't turn cars fast enough to keep the inventory fresh. The reconditioning centers didn't save enough money to pay for themselves either. Huizenga ended up closing them.

To funnel used cars into his growing number of lots, Huizenga tried buying Alamo and National car rental. That way he could grow his own inventory rather than compete with other dealers for late-model used cars. Once Alamo and National were through with a car, Huizenga figured, he could recondition it and resell it through AutoNation stores. But in practice, customers weren't wild about the plain-vanilla models favored by rental companies, and they certainly didn't want to buy nearly new cars when manufacturers were offering good deals on brand-new ones. "We made a lot of mistakes," admits Huizenga.

His latest strategy is built around a string of preexisting dealerships in the Denver area that Huizenga has acquired over the past several years. He renamed them all John Elway AutoNation USA, to capitalize on the godlike status of the retired Denver Broncos quarterback. (Huizenga bought Elway's six dealerships last year.) To promote the one-price concept, Huizenga blanketed the market with advertising and promotions. He stocked his lots with popular but hard-to-get models like the Honda Odyssey minivan. He put salespeople through up to 60 days of training so that they could learn how to sell a car without dickering, and he changed their commissions to reflect the number of sales they made rather than the gross profit from each one. Those who couldn't adjust to the new way of business, or who feared a cut in pay, moved on.

Since the change, Huizenga's market share has taken a huge leap forward, from 19% a year ago to 28% in June. But profit margins have suffered. The advertising is expensive, the cars were priced unrealistically low, and Huizenga's focus on new cars ignored used-car sales, which are actually much more profitable. During the quarter ended June 30, Huizenga says he made a pre-tax profit in Denver of just 1.9%, compared with 3% a year ago. That may be better than the average dealer, but those aren't the numbers he had in mind when he got into the car business in 1995. Nor do they appeal much to investors, who have driven down AutoNation stock from a high of almost $45 in 1996 to $12 recently. Huizenga says he hopes to raise margins to 3% by year-end, and he is determined to move them higher in the future. "If we thought we were going through this for 3% margins, we wouldn't have this conversation," he observes.

You have to believe Huizenga is serious this time. He has stepped back from his other ventures, which range from extended-stay hotels to the Miami Dolphins football team, to get more involved in the car business. He has restructured the company by divesting nonautomotive operations, changed the name from Republic Industries, and replaced several executives, including his partner of 19 years, Steve Berrard, with whom he co-founded the business.

In September, Huizenga named Mike Jackson president and CEO. Jackson, a respected industry veteran who used to be a dealer himself, ran Mercedes-Benz's North American sales and marketing operation. Instead of one-price selling, Jackson was installing what he called "credible" pricing by reducing dealer margins so much that it was uneconomical for them to dicker with prices any further.

Like Huizenga, one-time stockbroker Marshall Cogan had never before sold cars but figured he could apply skills from other businesses to reforming the auto dealer system. Beginning in 1990, he assembled 66 dealerships, more than a quarter of them Toyota/Lexus, into United Auto Group, the second-largest publicly owned retailer network. The idea was logical enough--with such scale, he figured, he could be more efficient than the competition at buying advertising, handling paperwork, and borrowing money. But the savings proved hard to find, and Cogan watched profits evaporate. After going public three years ago at $30 a share, UAG had fallen to $6 by earlier this year. In April, Cogan sold 40% of the company and relinquished operating control to Detroit industrialist Roger Penske, who runs several successful dealerships in California.

Austin Ligon, the president of CarMax, has also been struggling to create a 21st-century auto retailer. His six-year-old company is a chain of 32 used-car superstores and 21 new-car dealerships with one-price selling. "There is not some genetic code that makes you a 'car guy,' " he bristles. "A lot of people would have you believe that, but the question is really, Can you learn the business?"

Ligon is still learning, apparently. Some of the first stores he built were too big and didn't produce enough volume to support the advertising. Since then he's been developing smaller lots, clustering them in major markets, and attaching new-car dealerships to leverage customer traffic. But CarMax hasn't made an annual profit since it was founded in 1993, and Ligon expects it to do no better than break even this year.

If there is a genetic code for car guys, you would expect to find it imprinted in Dearborn, Mich. But Ford Motor's stab at creating a new auto-retailing system has run into its own set of problems. Ford's plan was to buy dealers in overcrowded markets, close some, and change the others to no-haggle selling. But the scheme went flat when experienced salespeople quit, corporate managers couldn't pick up the slack, and sales dropped precipitously. The Ford executive who ran the project quit abruptly in mid-September. "We went too fast, and we ran into things we didn't expect," concedes Ford sales and marketing boss Robert Rewey. Rewey remains a fan of one-price, which he says particularly appeals to younger buyers, but Ford has slowed down its expansion plans.

At Ford, as elsewhere, the one-price, no-haggle system is fundamental to reform strategies. The only problem is, it has never really worked. Inevitably, fixed prices put dealerships at a disadvantage to traditional competitors down the street, who--in an industry ruled by overcapacity--will always be willing to make a deal for $50 less. (Besides, the traditional dealer can always make it up by stiffing the customer on the trade-in.) Cynics say that no matter how much customers tell market researchers that they want pleasant surroundings and no haggling, the reality is that they prefer an atmosphere that resembles a rug bazaar. "People like service, but what they really want is low prices," says California marketing consultant Jeremy Anwyl. Only Saturn dealers have been able to make one-price selling stick, and then only because they enjoy exclusive territories and a well-loved product.

The Internet was supposed to be the magic elixir that solved auto-retailing problems, since it offered to dispense entirely with salesmen--plaid-jacketed, khaki-clad, or otherwise--and all the costs attendant on maintaining inventory and showrooms. However, the vision of buying a car with a few clicks of a computer mouse seems to be slipping further into the future. One key obstacle is regulatory: State franchise laws make it all but impossible for anyone but a dealer to sell a car to a customer. Even a GM employee in Detroit can't buy direct. Another major hurdle is psychological: Very few people want to fork over 50 large for a new Jaguar without having had a chance to feel the wood and sniff the leather.

As a result, services like Autobytel.com and Autoweb.com, which were supposed to do for cars what Amazon has done for books, have turned out to be merely a way for dealers to identify potential customers. Microsoft's CarPoint, which also started out as a referral service, has taken a step beyond that by forming a venture with Ford that would allow buyers to custom-order cars over the Internet. But they would still have to negotiate a price and take delivery from the same cigar-smoking salesman they were trying to avoid in the first place.

CarsDirect.com, another much-ballyhooed Internet service, which is partly financed by Michael Dell, looks less and less like a startling new departure and more like a traditional auto broker. CarsDirect will take an order to buy a car at a particular price, fill the order with a local dealer, and then deliver the vehicle. But it can't take orders for certain hot-selling models because dealers would rather keep that high-margin business for themselves. It sells other models for as much as $5,000 below invoice and admits to losing money on some sales--a familiar Internet strategy to gain market share. CEO and co-founder Scott Painter still says he expects to be profitable by 2001.

Another new wrinkle in Internet selling has been added by priceline.com, the heavily advertised site that allows users to bid for airplane tickets and hotel rooms. For cars, priceline lets customers specify the kind of car they want and make a reasonable offer--no $50 Porsches, please--then sends the bids out daily to nearby dealers. If the dealer doesn't like the price or doesn't have the car in stock, the sale doesn't get made. Priceline won't say what percentage of the orders it actually fills, but some dealers complain that its bargain-hunting customers can be prickly to handle.

Before the Web can have any hope of establishing a foothold, the state franchise laws have to be overthrown. Some experts predict that could happen in the next five years, and if it does, everything could be up for grabs. Websites could turn themselves into distributors by buying dealers, and manufacturers could sell cars directly to customers, while independent providers like Pep Boys or Jiffy Lube take over the job of providing service.

More likely, though, the traditional operator doing business in traditional ways will remain in place even then. For all the capital and brainpower that Huizenga, Cogan, Ligon, and others have put behind reform efforts, the old ways continue to thrive. Maybe that's because they are still best at serving the needs of manufacturer, customer, and retailer. To paraphrase Churchill's famous statement about democracy: The current system of selling cars may indeed be the worst possible--except for all the others that have been tried.