(FORTUNE Magazine) – The Empire State ballroom at the Grand Hyatt Hotel in midtown Manhattan typically serves as the setting for the kind of stuffy, state-of-a-business luncheons that are more snooze inducing than buzz producing. But as Amgen CEO Kevin Sharer stepped up to the podium on a Tuesday morning in late March, the scene in front of him had the feel of a rock concert. Roughly 500 investors and analysts had crowded in, jostling for position at the biotech bellwether's first-ever R&D Day and expecting to be wowed.

Sharer has done plenty to stoke their curiosity. Since taking charge in April 2000, he has focused on radically remaking the Thousand Oaks, Calif., company while attempting not to upset its cherished biotech roots, a tricky balancing act that has created new stresses even as it has fixed old problems. The makeover embraces every aspect of Amgen's business, and has ushered in new people, new partners, and a new approach to the product pipeline. During the past three years, for example, Sharer has recruited a host of execs from buttoned-down pharmaceutical firms like Merck and Eli Lilly. And one of the most prominent new hires, whip-smart R&D chief Roger Perlmutter, has borrowed from the Big Pharma playbook to install a "command and control" approach to drug discovery that seeks to marry the free-form creativity of biotechs with the more traditional, calculating approach of their older rivals in the drug business.

It's a new model--one that's being tested every day. But Amgen had no choice except to change. The biotech coasted through much of the 1990s, living off the nearly $17 billion in sales generated by two blockbuster drugs it had launched at the start of the decade, the protein-based injectable drugs Epogen (2003 sales of $2.4 billion) and Neupogen ($1.3 billion). With healthy profit margins, Amgen was converting plenty of those revenues into cash but very little of its R&D spending into the new products it needed to juice its easing growth.

Now Amgen's transformation from precocious biotech to full-blown pharma company is being watched carefully in all corners of a drug world that is struggling to regain its productivity mojo. While overall R&D spending has doubled over the past decade, the number of new molecular entities launched has dropped by a third, according to the Center for Medicines Research. Pharmaceutical firms like Merck and Pfizer have morphed into massive, staid marketing and sales operations, saddled with layers of bureaucracy and struggling to replace blockbuster drugs coming off patent protection. Biotechs, meanwhile, have suffered through a grueling three-year bear market as funding dried up and dozens of companies retrenched or dropped off the map entirely. In 1994, the T-shirt that biotech Millennium Pharmaceuticals distributed to all its employees boasted NOTHING IS IMPOSSIBLE. Seven years later, the shirt read FOCUSED EXECUTION. Not exactly a call to arms.

Over the past few years industry observers have pegged Amgen as possibly the drug industry's best hope for turning around this negative trend--especially as Sharer has aggressively embraced the challenge of running a company that's both big and creative. The high expectations are reflected in the fact that Amgen's market cap of $83 billion now exceeds all pharmaceutical firms save for giants Pfizer, Johnson & Johnson, and Merck. But with Amgen's ballooning size ($2.3 billion in profits on sales of $8.4 billion last year), analysts have become divided on whether its potential continues to justify a speculative biotech premium based on delivering earnings per share growth of 30% a year. Recent target prices for its stock--ranging from $62 to $90--reflect that confusion, and some analysts have recently ratcheted down their earnings projections.

All of which explains why the investors and analysts crowded into the Hyatt ballroom were so eager to hear what the Amgen CEO had to say about the pipeline. Expecting something big, they left a bit disappointed. Sharer didn't raise the earnings outlook and, more important, failed to surprise them with news of a new blockbuster set to hit the market. By the end of the day, Amgen's stock had fallen more than $2, to $57.83, off 20% from its July high of $72.

Days later when Sharer did come out with big news--announcing Amgen would purchase the 79% of cancer-focused biotech Tularik that it did not already own for $1.3 billion--the market basically shrugged. Still, Amgen's leadership remains confident in its ability both to defy the skeptics who are deriding it for becoming a card-carrying member of Big Pharma and to wow the critics who say it has lost its spark. "We're out to prove that an $80 billion company can still maintain that [biotech] mindset," says Perlmutter.

Kevin Sharer, 56, assumed the top spot at Amgen just as the bull market turned sour. He was new to the CEO suite but not new to Amgen, after serving as COO under former CEO Gordon Binder for eight years. During that time Sharer got a crash course in biotech. It was a subject he knew nothing about when Binder hired him in 1992 from MCI, where he served as head of the $6 billion business markets division but left after it became clear he wouldn't get the top job.

Highly ambitious, Sharer started his business career at McKinsey after leaving the Navy; he quickly moved over to General Electric in 1984, where the Iowa native served as a high-level advisor on M&A. In 1989, at the ripe age of 41, he received a plum assignment in GE's aircraft-engine business, but he quit on his first day and bolted to MCI. Sharer chalks up the decision to youthful hubris. 3M CEO Jim McNerney, who was working in GE's computer services unit at the time, remembers Sharer as "kinetic, bright--he's a drink from a fire hose," but prone to impulsive behavior.

Poor eyesight may have kept him from being a Navy pilot as his father was, but Sharer's vision was keen enough to see that times were changing as he moved into the top spot at Amgen. While Epogen and Neupogen had always enjoyed relative monopolies, Amgen in 2000 was readying products like Aranesp, a longer-lasting version of anemia fighter Epogen that would go head-to-head with Johnson & Johnson's blockbuster anemia drug Procrit. (Amgen patented Erythropoietin, the protein that stimulates red-blood-cell production, but a 1985 agreement between cash-poor Amgen and J&J divvied up the worldwide market.) Amgen was also planning to enter areas like inflammation and oncology, where competition was fierce. "We were going from a monopoly environment to a competitive environment, and I don't know any companies that have successfully done that--ever," Sharer says, before chuckling and adding, "so that caught my attention."

Sharer knew that Amgen's cherished corporate culture needed to be shaken up. Research (the lab work) and development (the clinical tests) had always been kept at arm's length at the company, where the two previous CEOs (George Rathmann and Gordon Binder) served as de facto R&D chiefs. And the marketing folks were not integrated into the drug development process whatsoever. Not coincidentally, Amgen hadn't managed to launch a new drug in almost a decade.

For Amgen to move forward, Sharer decided, that would have to change. Sharer had identified plenty of deadwood during his eight years inside the company. Finally in a position to call the shots, he began remaking the executive suite like a hired-gun called in from the outside. The new CEO purged the management ranks: Of the 40 VPs at Amgen when Sharer took charge, half are no longer there. In order to raise the competitive metabolism, Sharer made sure many of the new hires came from Big Pharma bastions, where cutthroat competition was the norm.

One of Sharer's first additions was George Morrow, 52, formerly head of GlaxoWellcome's $6.5 billion U.S. operations, who took over sales and marketing. When asked what was wrong with Amgen's commercialization process, Morrow smiles, leans in, and says, "How long do you have?" Rather than a flowing process, he says, it was a series of discrete steps. To integrate them, he quickly set up product strategy teams, co-led by a clinician and a senior marketing rep, who would "own the product."

Sharer's savviest hire, though, was Perlmutter, Amgen's first true head of research and development. During four years at Merck, where he rose to executive VP of research, and earlier as chair of the Immunology Department at the University of Washington, Perlmutter had earned a reputation in the scientific community that was without blemish. "People are drawn to him because of his raw mental horsepower and his knowledge of science," says Stephen Williams, head of biotech recruiting at search firm Bench International. At Merck, he led the team that developed the anti-inflammatory Vioxx, now a $2.5 billion drug. Like Morrow, Perlmutter, 51, encountered a silo environment, with research and clinical development each doing its own thing. When Perlmutter brought the senior members of the two groups together in April 2001, he says, it was the first time many had sat in the same room. (Daniel Vapnek, a former head of research at Amgen who left the company in 1997, disputes that, saying that during his tenure there was "good communication between research and development.")

When R finally sat down with D, it became obvious to Perlmutter that quite a few of Amgen's research projects would never see the light of day. Amazingly, there were more active research projects at Amgen than at Merck. So Perlmutter axed about half of them and funneled R&D funds into the most attractive compounds. "What I wanted to do was introduce into Amgen the rigor and discipline that characterizes the very best R&D organizations without destroying this entrepreneurial spirit," Perlmutter says. "To do that, you need to introduce command and control."

Perlmutter set about assembling a staff comfortable with rigor and discipline. At the top he installed two familiar faces from his Merck days: Joseph Miletich as senior VP of research and preclinical development; and Beth Seidenberg as senior VP of development. Both sit on Amgen's executive committee, which means that four of Amgen's top 11 officers are former Mercksters. (The other is CIO Hassan Dayem.) Some might question whether hiring so many people from a struggling company like Merck was a smart move, but Merck has never had trouble attracting talented scientists--retaining them is another matter. Moving down the pyramid, Amgen's business units are brimming with Big Pharma hires.

This rapid assimilation of new talent--60% of the R&D staff has arrived since January 2001--has had a profound impact on Amgen's identity. Mark Levin, CEO of Millennium Pharmaceuticals, a biotech based in Cambridge, Mass., that grew rapidly during the late 1990s, says that he often recruits former Big Pharma staff, but he prefers it if the candidate has spent time at another biotech first, as a sort of deprogramming period. Biotech recruiters say that's often not necessary at Amgen, because its culture now has more in common with Big Pharma than with a small biotech. "It's easier to go from Glaxo to Amgen than from a biotech startup to Amgen," says John Phillip, a biotech headhunter. And however necessary it was, Perlmutter's more rigid approach to R&D has rankled some employees. Says an Amgen staffer who came from a large pharmaceutical firm and works in preclinical development: "There used to be a lot more weight put on the science. Now everything is a business decision."

New employees are often surprised at the many layers of management at Amgen. Victor Fung, associate director of process development at Amgen's snazzy new $625 million R&D center overlooking Seattle's Elliott Bay, says so many people are involved in decisions that "I feel like a switchboard operator just making connections." The preclinical employee adds, "It's completely different from the company I came to work for."

Sharer says he takes "great exception" to the claim that business considerations trump good science. And although both he and Perlmutter have heard the lament about stifling bureaucracy, they say it's an unfortunate reality in a large organization, especially one growing so quickly. As Amgen's culture digests all the change, the hiring door is swinging both ways. William Robbins, a venture capitalist based in Los Angeles, says he sees "some departures at Amgen we hadn't seen before." Turnover at Amgen, which stood at 5.2% in 2000, is now 6.7%, slightly up but still below what you'd find at large pharmaceutical company. So while they might not love the new Amgen--or "Merck West," as some derisively call it--members of the old guard are learning to live with it.

Amgen's location has always been a blessing and a curse, situated just north of Malibu in the Santa Monica Mountains. Those of you living in less idyllic climates can easily imagine the blessings. The curse is that Thousand Oaks, northwest of Los Angeles, is a dead zone for biotech. (In fact, Amgen's hometown used to be so desolate that episodes of Gunsmoke were filmed there.) That isolation has shaped Amgen's identity. While Amgen's biotech brethren like Genentech and Genzyme have long partnered with thriving startup communities in their respective hometowns, Amgen did little, content to sit on the piles of cash generated by Epogen and Neupogen.

Over time, Amgen earned a reputation as a bit aloof, even arrogant, in its dealings with smaller biotech companies that were eager to trade their discoveries for some of Amgen's cash and cache. "Amgen traditionally has been thought of as difficult to approach," says Cynthia Robbins-Roth, a biotech consultant. "They had two successful products, so they didn't seem to feel the need to pull in new technology." (Amgen's third blockbuster product, rheumatoid arthritis drug Enbrel, came with its $10.3 billion acquisition of Immunex in 2002.)

When Sharer took over, he quickly decided the company had to ditch its go-it-alone strategy and aggressively seek out partnerships to bolster its pipeline. However, Amgen's new brain trust did not realize they might have a problem closing those deals until they lost the bidding for cardiovascular specialists Scios, which Johnson & Johnson bought this past February. When Perlmutter met with Scios, he learned that its people had an outdated view of Amgen, and it suddenly hit him: "Nobody out there knows what a different company this is. We'd better get out and talk to people." So they hit the road, giving presentations to startups in Boston, San Francisco, San Diego, and Seattle. The payoff has been quick: Amgen signed 68 deals last year, compared with 40 in 2002 and 25 in 2001.

One deal that typifies the new approach is the partnership Amgen formed last year with South San Francisco oncology specialist Tularik, the biotech that Sharer recently agreed to acquire. Oncology is the hottest area of drug development in biotech, and Amgen is plunging more resources into cancer than any other therapeutic specialty. And rather than focus on medications that play a peripheral role, like fighting infections in chemo patients, Amgen is now developing drugs that battle cancer itself. Amgen had been talking to Tularik about a collaboration for years, but nothing happened until Perlmutter brought a team up to the Bay Area in February 2003.

After the meeting Perlmutter spoke privately with Tularik CEO Dave Goeddel, a biotech pioneer who in 1978 was the first scientist hired by Genentech, where he worked on insulin, biotech's first drug. Tularik was already in advanced discussions with two companies for its ocotogene (cancer gene) discovery program. "That's not acceptable," Perlmutter said. "We're the ideal partner for you." In less than a month Amgen made a sweeter offer and sealed the deal. "The advantage of being a small company is we can act on decisions," says Goeddel. "To have a large partner who can act with that same speed is great."

The Tularik deal also displayed Amgen's drive into small-molecule drugs, the type that can be made into pills and have heretofore been the exclusive domain of Big Pharma. In early March, Amgen received FDA approval for its first oral medication, Sensipar, which treats a gland disorder called secondary hyperparathyroidism. Amgen's move into small molecules is significant for two reasons. One, it increases the numbers of ways it can attack a particular disease. Two, small-molecule drug development is a much different animal than large-molecule research, requiring a huge investment in new staff and facilities. Sharer believes that this commitment is necessary to safeguard Amgen's future.

Amgen's R&D Day in New York was its first such presentation because its development pipeline wasn't much to brag about before. Aranesp (the longer-lasting version of Epogen) and Neulasta (an improved successor to Neupogen released in 2002) have been wildly successful, but both are basically variations on a theme. Meanwhile, Amgen has broadened the use of its rheumatoid arthritis drug Enbrel, which reached blockbuster status ($1.3 billion in sales) last year, to treat ankylosing spondylitis (a rare arthritis of the spine), psoriatic arthritis (suffered by 23% of the 4.5 million American adults who have the skin disorder psoriasis), and, later this year, the company hopes, psoriasis itself. But Sensipar is the only drug it has launched in two years. Rival Genentech, meanwhile, has scored some high-profile approvals recently, such as the colon-cancer drug Avastin, and last year brought almost 20 programs into the pipeline--all with an R&D budget of less than half of Amgen's $1.6 billion. "Considering the size of [Amgen], they haven't put a lot forward," says Burt Adelman, head of development at biotech rival BiogenIdec.

That should change. In Perlmutter's first three years (2001--03), Amgen put 23 programs into development (meaning it started the preclinical process), which is one more than it had put forward in the previous decade. The most promising is AMG 162, a monoclonal antibody that targets osteoporosis and will enter Phase III clinical trials this year. (The Tularik acquisition will also add five clinical programs to the pipeline.) Amgen should enroll nearly 50,000 patients in clinical trials this year, up from 27,700 in 2002. "Compared with many large pharma companies, we are enormously productive," says Perlmutter.

The comment goes to the heart of Amgen's dual identity. Amgen execs want to have it both ways--they enjoy comparing themselves to Big Pharma when the comparison makes them look good, then flatly refuse to be compared to Big Pharma when the subject is, say, culture. It's understandable that Amgen, or any biotech, might reel in horror at being branded Big Pharma--the label bears the stigma of creative atrophy. But Amgen should just accept it and move on. "I think of Amgen as a pharmaceutical company now," says the Amgen preclinical employee. "But it had to happen for the company to be successful."

Visitors to Sharer's office can't miss the huge, modern art portrait of General George A. Custer staring ominously down from the wall, which reminds Sharer to never underestimate his opponents. Last year, after reading Edgar Vincent's biography of Horatio Nelson, Sharer found a new role model and commissioned a portrait of the English naval hero. Extremely ambitious and self-confident, Nelson is famous for taking huge risks in defiance of orders, but, unlike Custer, he didn't rush headlong into peril. Rather, Vincent argues, Nelson was actually a skilled communicator and collaborator, capable of playing office politics to generate internal support for his innovative plans. In a way, that's exactly what Sharer hopes to do at Amgen--act boldly and swiftly but with consensus, taking calculated risks that have a big payoff. The rest of the drug world will be watching.