(FORTUNE Magazine) – Listening to Ray Gilmartin, you can't help but be struck by how matter-of-fact and calm the man seems. The 63-year-old Merck chairman and CEO is soft-spoken, quick with a smile, and unfailingly polite, no matter how pointed the question he is asked. Interviewing him, as FORTUNE did at length recently, you'd hardly guess that he is facing the most serious crisis of his ten-year tenure atop the storied $22-billion-a-year pharma giant.

But he is. The event precipitating the crisis, of course, was Merck's announcement on Sept. 30 that it was recalling Vioxx, its $2.5-billion- a-year arthritis medicine, because the drug has been shown to double the risk of heart attacks and strokes in long-term users. On the day of the announcement, Merck suffered its own heart attack as investors fled the stock, and $27 billion in market cap--27% of the total--evaporated. (The company has subsequently lost an additional $6 billion in market cap.) Merck has taken a devastating public relations hit too; the recall has raised serious ethical questions about whether Merck inappropriately downplayed Vioxx safety issues over the past several years. And then there's the litigation problem: Although the company won't disclose how many lawsuits have been filed so far, roughly 20 million Americans had taken Vioxx by the time of the recall. According to an estimate in the Oct. 21 issue of the New England Journal of Medicine, tens of thousands of them may have had "major adverse events" attributable to the drug. Tort lawyers around the country have already set up toll-free numbers to solicit potential clients among Vioxx patients. The cost of settling the lawsuits will almost surely be in the billions of dollars, quite possibly in the tens of billions (see box).

To which Gilmartin's fundamental response is: Stay the course. When he got the bad news about Vioxx from research chief Peter Kim, he says, "my reaction ... was that we were going to make a decision about Vioxx totally in the interest of patient safety." That, he says, is what the company did. Those who believe Merck should have removed the drug from the market years earlier are wrong, he insists--a Merck-sponsored study showing the increased risk of heart attacks (after 18 months of taking the drug) was, in the company's view, the first concrete evidence of a serious problem.

As for the effect of Vioxx on Merck's earnings, Gilmartin concedes that withdrawing the blockbuster will hurt short-term profits. But he insists that the company has the financial wherewithal to weather the storm and--with a little help from speeded-up cost cutting--get earnings back on track. "The thing to keep in perspective is how financially strong the company is," he says. "We're fortunate to have been managed conservatively, because this is the kind of event that you want to be able to protect yourself against." Even after the Vioxx recall, both Standard & Poor's and Moody's kept Merck's triple-A bond rating (though both agencies say the ratings are under review). The company has, at last count, more than $10 billion in liquid assets.

If Vioxx were Merck's only problem--if it were a singular, aberrational event, lacking any larger context--Gilmartin might well be right. Merck could take its short-term earnings hit, settle its lawsuits, and move on. But it's not. Instead, Vioxx is only the latest in a series of problems at a company that, not so long ago, seemingly could do no wrong. Last November, Merck was forced to cancel work on major new drugs for depression and diabetes that its executives had long touted as blockbusters in the making. The depression drug didn't work in a pivotal clinical trial, and the diabetes drug was found in animal studies to pose a risk of cancer. Meanwhile its anticholesterol blockbuster, Zocor, which accounts for $5 billion in annual revenues, is going off patent in 2006, and Merck urgently needs to replace those sales. Its earnings, though still a monumental $6.8 billion last year, have been in decline since 2001. Since 2000, when the stock peaked at $95 a share, Merck has seen its market cap erode by a staggering $150 billion. Once the world's largest drug company, Merck is now No. 6. Failure haunts it like never before.

And that's the real crisis at Merck. At a time when the entire pharmaceutical industry is struggling, no company is struggling more, a fact both shocking and sad in light of Merck's noble past. Its researchers have racked up decades of dazzling successes, including major new medicines to treat hypertension, osteoporosis, high cholesterol, and AIDS. The company has donated millions of dollars worth of medicine to fight infectious Third World diseases such as river blindness, continuing a tradition of social responsibility that dates from the 1940s. Among businesspeople, Merck has long been revered: in FORTUNE's corporate reputation survey it was the Most Admired Company in American business for seven straight years in the 1980s. So too on Wall Street: As far back as most investors can remember, Merck has ranked among the bluest of blue-chip stocks.

Look a little more closely, then, and you can see the strain on Gilmartin's face. Before the Vioxx crisis hit, the board was lining up executive recruiters to begin a methodical search to replace him when he reaches retirement age in 18 months. That search has now taken on new urgency. While several key directors told FORTUNE that the board continues to back Gilmartin, it seems clear that the next CEO will come from outside the company. The board has realized that Merck needs a new leader who can see the business with fresh eyes.

Wall Street, meanwhile, has become impatient. "I wish I could give you my Rolodex," says Richard Evans, the pharma analyst for Sanford C. Bernstein. "My clients, the guys who move health-care stocks, are pissed off. No doubt about it."

"Merck is living in the past," adds Barbara Ryan, an analyst who covers the company for Deutsche Bank. She's mad about the drop in the stock, sure, but there's some truth in her complaint. For as glorious as Merck's past has been, that's also where the seeds were planted that grew into its troubles today.

A little more than a week after pulling the plug on Vioxx, Gilmartin and Kim were on a plane to Boston to dedicate Merck's newest research lab, a multimillion-dollar structure near the Harvard Medical School. Built to house R&D for potentially lucrative drugs to treat Alzheimer's, cancer, and obesity, the gleaming glass and metal edifice is as imposing and steely as the man whose name it bears: Edward M. Scolnick, physician, world-class scientist, and ex-president of Merck Research Labs. Not every corporate chief scientist gets a building named after him, but Scolnick was much more than that. He presided over an unprecedented burst of R&D productivity--in a five-year period in the 1990s, Merck launched no fewer than 15 unique drugs, many of which became blockbusters. What's more, for at least a decade before he retired in 2003, Scolnick was Merck's de facto No. 2. And for the last six of those years, he sat on the board--the only inside director besides the CEO.

Scolnick is remembered by many Merck researchers as the sharpest and most abrasive intellect they've ever encountered. (Now a member of the teaching staffs at Harvard and MIT, Scolnick declined to be interviewed.) It's the sort of praise that's usually easy to discount, except that Merck's scientists are themselves among the world's brightest. The perfectionism that propelled the tweedy, tall, Boston-born Scolnick through Harvard Medical School--and punctuated the roughly 200 scientific articles he authored--also suffused his leadership at Merck. "You never went before him unprepared," says a former Merck cancer researcher. "He would begin probing very directly and very quickly. He would often identify some controlled experiment that should have been done and wasn't."

The drive for perfection permeated Merck's research organization. Scolnick's leadership fostered among Merck's researchers the conviction that they were the best in the business. The sentiment was probably justified; in the past two decades, for instance, Merck has published more scientific papers and patented more compounds than any of its competitors. It created a culture of scientific insularity and hubris that dogs the company today, but for a long time, Merck's vaunted reputation in science gave the company a quantifiable edge in the industry and served shareholders very well. Merck was known for producing fastidious supporting documents for its FDA approvals, which came through faster than any of its competitors'. Between 1995 and 2001, for instance, Merck presented 13 major new drugs. All were approved, with an average review time of less than 11 months. (Vioxx won approval after just six months of review.) By contrast, Pfizer's submissions during the same period faced an average review time of more than two years.

Approval times are important. A quick nod from the FDA can mean a year or more of additional time on the market before a drug loses its patent--potentially worth hundreds of millions in sales. Merrill Lynch estimates that Merck gained some $3.3 billion of extra sales between 1995 and 2001 as a result of speedy approvals.

Scolnick served under two CEOs. He joined Merck under the legendary leader Roy Vagelos, who reigned for nine years. But Scolnick's power crystallized with Gilmartin's arrival in 1994. Vagelos had made a number of acquisitions--efforts to diversify Merck from its core drug-research business. The board decided to bring in a successor who could return Merck to being a more focused, research-driven drug company. Their pick was Gilmartin, who in 1994 became the first nonscientist ever to run Merck. His training had been in electrical engineering, but as former CEO of Becton Dickinson, he arrived with a reputation as both a turnaround artist and a wizard at efficiency.

Gilmartin succeeded in selling off many of Merck's noncore businesses, including eventually Medco, the $30-plus-billion-a-year drug-distribution company it had purchased just before Vagelos retired. Meanwhile Scolnick was generating phenomenal growth. The drugs in the pipeline were his babies. He knew them intimately and defended them to upper management with the same ferocity he applied to managing scientists. Gilmartin deferred to Scolnick on matters of science--and much more. "When Ed developed a compound, he had strong ideas about everything from its medical utility to its positioning in the marketplace," says Douglas Greene, former executive vice president of clinical science and product development.

The most dramatic example of Scolnick's vision was Merck's positioning of Fosamax, its innovative anti-osteoporosis medicine, in 1995. Osteoporosis--the bone loss women often experience in old age--attracted little public attention at the time. Scolnick, however, envisioned a market that extended far beyond ailing old ladies. At his urging, Merck launched an osteoporosis public-awareness campaign in the year leading up to Fosamax's launch. The drug generated $280 million in sales in its first full year on the market (and $2.7 billion in sales last year).

But Scolnick's stewardship of Fosamax didn't end there. Soon after its release, the FDA threatened to revoke its approval when researchers discovered that Fosamax caused some patients to experience erosion in the esophagus. "Ed was completely transparent about the issue," says Greene. "He wrote to doctors, in his own hand, explaining the causes. He pushed through every bit of data with the FDA." In the end the FDA agreed to let Merck keep Fosamax on the market, albeit with a warning label that told patients to sit upright for an hour after taking the drug. "Ed saved the drug. He literally saved it," says Greene.

Scolnick saved Gilmartin too. Wall Street had been wringing its hands over Merck's future when Vagelos retired. Analysts knew Merck would ultimately face patent expirations on Mevacor, Pepcid, Prilosec, and other major drugs, which together accounted for annual sales of more than $5 billion. Though the analysts were familiar with Merck's pipeline, they didn't fully appreciate the promise of the drugs that Gilmartin and Scolnick insisted were future winners.

But within 18 months of Gilmartin's arrival, Merck began its spectacular run of blockbusters. In 1995 there were Fosamax and Cozaar, a $2.5-billion-a-year hypertension treatment. The following year came Crixivan for HIV. And in 1998, Merck's labs produced five medicines, including the $2-billion-a-year asthma remedy, Singulair, and Propecia for baldness. The blockbusters helped boost Merck's share price by 188% between 1995 and mid-1999. Gilmartin and Scolnick shared the credit; in 1997 the board of directors elected Scolnick as one of its members "in recognition of his extraordinary work," says director William Bowen.

By 1999, however, Scolnick was beginning to plan for retirement and Merck was facing yet another spate of patent expirations. This time there wasn't a slew of miracle drugs in the pipeline. Although it is absurd to say that discovering blockbusters is ever easy, it is certainly true that by then drugmakers had plucked much of the low-hanging fruit. Classes of drugs such as statins, which reduce cholesterol by tweaking a liver enzyme well understood by scientists, had turned into some of the biggest blockbusters of all time. But as researchers pursued novel therapies for diseases such as cancer, arthritis, and diabetes, whose molecular underpinnings are far less understood, it became clear that developing a new generation of blockbusters was going to involve far more difficult science.

Merck wasn't alone in facing this conundrum; so was every big pharma company. Some, such as Pfizer, began going outside their labs to purchase companies that had either promising pipelines or drugs already on the market. (Pfizer got its hands on Lipitor by buying Warner-Lambert in 2000 for $120 billion.) Other companies raced to market with lucrative "me too" drugs such as the antidepressant Zoloft that mimicked established blockbusters.

But that wasn't the Merck way--nor Scolnick's. No, Merck would discover new drugs the way it always had: by doing its own novel, groundbreaking science. Yes, the risks were higher, but Merck felt they were worth taking. Several of the molecules Scolnick's teams were then working on--particularly two drugs for diabetes and depression--had tremendous potential payoffs. But they also had potential problems, and if the drugs failed, Merck would have little else almost ready for the market.

Scolnick, meanwhile, was also trying to choose his successor. His favorite candidate was Roger Perlmutter, a renowned immunologist and charismatic manager he had recruited in 1997. The pressure on Scolnick became clear to those working most closely with him. "Ed alienated many researchers by becoming more abusive in meetings, playing favorites, and making succession choices that didn't make sense to people," says Eve Slater, former vice president of clinical and regulatory development, and a 19-year veteran of Merck's labs. "The result was, people were afraid to tell him where the problems were. There was suddenly this emperor's-new-clothes mentality."

By the end of 2000, Perlmutter found himself on the outs with Scolnick. He soon left to become head of R&D at Amgen, where he helped recruit key researchers and executives away from Merck's labs. A few weeks later Scolnick hired another heir apparent: Peter Kim, a famed biomedical researcher from MIT. Embracing Merck tradition, Kim proclaimed his desire to have Merck scientists go it alone, focusing on novel drugs and avoiding the me-too stuff. "I want to work on things which nobody's done before, because that's what turns me on scientifically," he told FORTUNE in 2001, shortly after joining the company. "I don't want to be developing the next antihistamine."

When Scolnick retired in 2003, Kim officially took the helm of Merck's research. But the Scolnick vacuum has so far proved hard to fill. The labs have been in a funk, as a number of those big bets Merck and Scolnick had made in the late 1990s failed. Last year Kim had to discontinue research on four potential blockbusters--including the diabetes and depression drugs, which were in big, costly Phase III trials. Financially, last year was one of Merck's worst. Sales grew just 5%, and net income fell for a second consecutive year, dropping 4.5%. Between 2001 and the end of 2003, Merck's shares lost nearly half their value. Gilmartin was forced to trim the workforce by 7%. Wall Street began calling for his ouster.

And then came Vioxx.

Vioxx was the last of the Scolnick blockbusters. Discovered in a Merck lab in 1994, the drug was one of a new class of painkillers called COX-2 inhibitors, which reduce pain and inflammation without the side effects--primarily ulcers and gastrointestinal bleeding--that such common painkillers as ibuprofen can cause. The latter drugs are called nonsteroidal anti-inflammatory drugs (NSAIDs); by some estimates, intestinal bleeding induced by NSAIDs kills more than 10,000 Americans a year.

Vioxx worked beautifully in clinical trials with arthritis patients, and the FDA duly approved it for marketing in May 1999. Scolnick was ecstatic--it was surely going to be a multibillion-dollar drug. Scolnick made it known to the press that he was taking Vioxx himself for back pain. On the cover of its 1999 annual report, Merck boasted that the drug was its "biggest, fastest, and best launch ever."

Over the next five years Vioxx also became one of the great triumphs of direct-to-consumer marketing. In the time the drug was on the market, Merck spent more than $500 million on commercials trumpeting its virtues, according to the health-care industry research firm Verispan. By the time it was withdrawn, some 20 million Americans had taken it, and Vioxx was generating $2.5 billion in annual sales, making it the second-best-selling COX-2 inhibitor. (No. 1 was Celebrex, a $3-billion-a-year drug that Pfizer acquired when it bought Pharmacia last year.)

But there was always a cloud hanging over Vioxx. Even before the FDA had approved the compound, researchers outside Merck had found evidence that it--and other COX-2 inhibitors--might increase the risk of a heart attack. In 1998 a group at the University of Pennsylvania discovered that COX-2 inhibitors interfere with enzymes thought to play key roles in warding off cardiovascular disease. "We immediately postulated to the companies developing COX-2 inhibitors, and subsequently reported in print, that this was something that could lead to heart attacks and strokes," says Garret A. FitzGerald, the senior author of the study.

Merck says, however, that right up until it withdrew Vioxx, the evidence of cardiovascular effects was inconclusive, if not downright conflicting. For instance, patients who took Vioxx in the company's first round of clinical trials had about the same rate of heart attacks and strokes as patients who took NSAIDs or a sugar-pill placebo.

Then, in early 2000, Merck reported startling results from a trial it had initiated in 1999 to show that Vioxx posed less gastrointestinal risk than NSAIDs. Arthritis patients taking Vioxx had three times as many serious cardiovascular events as those on naproxen, an NSAID that is sold under the Aleve brand, among others. But as Peter Kim would later explain, these data did not necessarily mean that Vioxx increased the risk of heart attack. For one thing, there were studies suggesting that naproxen lowered heart-attack risk--so maybe that's what was happening: Naproxen was making Vioxx look bad by comparison.

Still, in April 2002, Merck updated the Vioxx label to include information about the possible cardiovascular risk. And even before then, Merck had decided it needed to investigate more closely the possible link between Vioxx and heart trouble. Just before Merck executives were learning of the results of the Vioxx-naproxen trial, the company embarked on a long-term study, called APPROVe, to see whether Vioxx would lead to a reduction of colon polyps. Merck decided to use that study--a true controlled trial that compared Vioxx with a placebo instead of another drug--to test definitively whether the arthritis drug increased cardiovascular risk.

On Sept. 23, 2004, a Thursday afternoon more than 4½ years later, Kim called Gilmartin at the office and told him that the APPROVe study was indeed showing that patients using Vioxx had a demonstrably higher rate of heart attack--but only after 18 months of regular use. "Peter told me that the outside investigators had called and suggested we end the study because of cardiovascular concerns," says Gilmartin.

To Gilmartin and other executives, the decision to recall Vioxx is nothing if not a sterling example of Merck hewing to its ethical traditions. After all, Merck made no attempt to suppress information about Vioxx; after it saw the results of the Vioxx-naproxen trial, for instance, it immediately put out a press release and informed the FDA. It added warning language to its Vioxx label. And it voluntarily conducted the research that ultimately uncovered the problem.

Gilmartin says that when Kim came into his office the following Monday and told him he thought Vioxx should be recalled, the research chief also pointed out that Merck was under no obligation to do anything of the sort. The company could simply go to the FDA "and have the prescribing information for the product updated with these new findings," Gilmartin says. Indeed, the majority of outside clinicians Merck consulted over the next few days suggested it do just that, since there were clearly millions of people who were benefiting from Vioxx and not getting heart attacks.

But Gilmartin told FORTUNE he never had any doubt about his course of action: "Withdrawing the drug was going to be the responsible thing to do." Patients using Vioxx could switch to other COX-2 inhibitors, such as Celebrex, or to NSAIDs. He adds, "It's built into the principles of the company to think in this fashion. That's why the management team came to such an easy conclusion."

Most rank-and-file employees feel the same way, says Kim. "This is a blow to the company," he concedes. But, he adds, there has been "an incredible outpouring of emotion that says, 'I'm proud that we did the right thing. And I'm proud to be part of an organization that would actually do the right thing.'"

Not everyone agrees--to say the least. Since the withdrawal, Merck has been subjected to a torrent of criticism from scientists who believe that the warning signs were plentiful--and that Merck largely ignored them because it was so hungry for sales. "If Merck were truly acting in the interest of the public, of course they should have done more studies" on Vioxx's safety when doubts about it first surfaced, says Harvard researcher Daniel H. Solomon, who has studied COX-2 inhibitors' side effects. "But that's not the way businesses operate. And the FDA seemed to really drop the ball on Vioxx," he says, by not pressuring Merck to resolve the doubts faster.

William Castelli, former director of the Framingham Heart Study, an influential investigation of cardiac risk factors, raises another issue. Since COX-2 inhibitors reduce inflammation, a probable risk factor for heart disease, many researchers expected Vioxx to reduce rather than to raise the risk of heart attacks. When some studies "suggested the exact opposite," he says, "it should have rung everyone's bell that something was not right."

Merck's most vocal critic is Eric J. Topol, chairman of cardiology at the Cleveland Clinic. Shortly after Vioxx was withdrawn, he wrote a stinging rebuke in the New England Journal of Medicine. "Had the company not valued sales over safety," he wrote, "a suitable trial could have been initiated rapidly [to pin down Vioxx's cardiovascular risk] at a fraction of the cost of Merck's direct-to-consumer advertising campaign." In 2001, Topol had been one of a number of researchers who raised questions about Vioxx in medical journals. He also says that at about the same time he made several phone calls to Gilmartin's office, offering to visit the company and present his detailed analysis of the risk. "He never responded to one call," adds Topol. (Kim acknowledges that Merck did not talk to Topol, though he does say the company spoke to FitzGerald, the Penn scientist whose 1998 study suggested that all COX-2 inhibitors could pose heart risks.)

Still, if Merck is right that Vioxx doesn't elevate cardiovascular risk until it is taken for 18 months--an effect that requires a multiyear trial to reveal--is it really fair to expect the company to have nailed down the risk quickly? Topol claims it is; he says that if the company had been primarily concerned with investigating Vioxx's safety, it "could have had the answer years ago" by testing the drug specifically in heart patients rather than in a broader population. Such a trial might have revealed Vioxx's cardiovascular dangers in months rather than years, he says.

The notion that Merck did not act as quickly or as ethically as it could have has the potential to haunt the company for years to come, as Vioxx lawsuits pile up and trials get under way. Even in the short term, the stain on Merck's image is taking a toll. "People are not looking at Merck in the same light as they used to" because of Vioxx, says Stelios Papadopoulous, vice chairman of SG Cowen Securities and a veteran pharma analyst. "There was a lot of debate about its side effects over the past few years, so now many people can point fingers and say, 'I told you so.' " "Merck may have held the moral high ground ten years ago," says Sanford Bernstein's Richard Evans, "but not anymore."

For decades Merck's stellar reputation helped it recruit the best scientists. It helped get the company's drugs swift approval by the FDA. It caused investors to give the company the benefit of the doubt. Merck's legacy lasted a long time. After the Vioxx debacle, however, fewer people are calling Merck the gold standard. Vioxx has reduced Merck to just another struggling pharma giant.

Even as Gilmartin and the board deal with the Vioxx fallout, they must turn their attention to the future. They face a quandary: How will Merck become a growth company again--and who will lead it? Although Merck has a number of impressive medicines all slated for release over the next few years--treatments for rotavirus, human papilloma virus, shingles, and diabetes--those drugs, according to most analysts, won't make up for the loss of Vioxx and the impending expiration of Zocor's patent. Added to that misery is the possibility that Merck's next big drug, another COX-2 inhibitor called Arcoxia, will be delayed at the FDA or, worse, rejected. (As FORTUNE went to press, Pfizer warned doctors that Bextra, one of its COX-2 drugs, might jeopardize the health of coronary-bypass patients.)

Wall Street has its own prescription for Merck. It thinks the company should merge with Schering-Plough. Schering and Merck seem to be working well together on a lucrative joint venture that recently launched Zetia and Vytorin, cholesterol-lowering medicines. A merger would enable Merck to reap the efficiencies of combining operations and garner the full projected sales of the franchise, which are expected to peak at $4 billion a year, rather than share them with Schering. Merck would also gain Schering's CEO--turnaround specialist Fred Hassan--who could then succeed Gilmartin.

Whatever path Merck takes, it almost certainly won't be that tidy. Gilmartin is dead-set against a large-scale merger. "The prevailing attitude in the industry is that mergers have not worked and they don't make sense," he points out. "It's very clear that increased size has not led to research productivity. A large-scale merger does not add to our pipeline, nor does it add to our long-term growth."

In Gilmartin's view, Merck needs no shift in strategy: It's already on the right track with a plan he began implementing some time ago. It involves reducing Merck's costs while entering partnerships with smaller, innovative companies and licensing promising compounds. Gilmartin admits that Merck hasn't always played well with others, but he insists the company is now stamping out the haughty go-it-alone attitude in the labs. His evidence: In 1999, Merck completed ten licensing deals. Last year the company signed 47 partnerships. And so far this year it has completed 41 such deals. "The theme is 'The doors are open,' " says Gilmartin.

But Wall Street views most of these deals as too little, too late--and at least one important board member suggests that Merck may need to rethink its strategy. Former Honeywell CEO Larry Bossidy told FORTUNE that at Merck's next board meeting, "We're going to take a look at the industry and Merck's future in it. What's the operating mode now?" In the past the board never even discussed the prospect of Merck doing a large deal, he says, because "frankly there's been no need for that." Now, however, "in light of these new realities, will it stay off the table? That hasn't been determined yet."

Most of all, what the board needs to face up to is that the Merck way, which brought the company such greatness for so long, is no longer the one true path. On the contrary, the risks from staying on that path will only rise. Merck can become a great company again--but only if it is willing to move beyond its glorious past. ■