(FORTUNE Magazine) – In business, as in life, there's no such thing as a perfect plan. So you can imagine how Eli Lilly's chief executive, Sidney Taurel, feels as he surveys the jagged urban terrain of Indianapolis from his office window, having orchestrated his company's mostly successful transition into the post-Prozac era. Born a Spanish citizen in Casablanca, Morocco, and schooled in France, Taurel, 55, is looking decidedly Midwestern today, dressed in a mustard button-down shirt and khakis. Still, he's oozing Continental flair--and confidence. "I'm sure we've come out of this challenging period a much stronger company," he says.

Indeed, Lilly is a bright spot in Big Pharma, widely acknowledged to have the industry's most bountiful pipeline of new products. Since May of 2003, Lilly's stock has outperformed those of its peers. The company's shares posted a total return of 27%, compared with the Amex pharmaceutical index's meager 5.8% gain over the same period. Just weeks ago Citigroup Smith Barney analyst George Grofik upgraded Lilly's shares from hold to buy, telling investors the company's next wave of new drugs "may be underappreciated by the market." Lilly's recovery is all the more impressive considering that, with $12.6 billion in sales last year, it is one of beleaguered Big Pharma's smaller companies. But even with all of Taurel's success, his plan to revitalize Lilly has come with sacrifices and unintended consequences that could hinder its growth. To fund the basic science that refilled its pipeline, the company may have cut corners in manufacturing and marketed existing products so aggressively that it broke laws.

To the outside world, Lilly's future appeared bleak just a few years ago. In 2001 the company's premier, $2.6-billion-a-year drug, the iconic antidepressant Prozac, lost its patent protection--two years sooner than expected--after a courtroom battle against generic drugmakers. Aside from being a major influence in the way the medical world treated and even perceived depression, the drug had powered Lilly's meteoric sales growth for more than a decade. In 2000, Prozac accounted for a quarter of the company's $10.8 billion in revenues. But by the closing months of 2001, as the drug battled with new generic antidepressants, Prozac's quarterly sales had dropped 66% from the previous year.

Without Prozac to lean on, Lilly fell into a funk. Around the same time the company started having manufacturing problems. Its factories received numerous citations from the FDA for violations, and Lilly's efforts to address them slowed pill-making and product launches. Wall Street soured on the company--Lilly's shares lost nearly a third of their value in the days following the U.S. Court of Appeals patent ruling in August 2000. Even as the '90s bull market collapsed, the pharma industry hit a dry spell in drug discovery; few analysts gave Lilly a fighting chance of replacing Prozac's blockbuster revenues.

What analysts didn't appreciate was that Taurel had a plan. He and Lilly's upper management had prepared for the day when Prozac's reign would end. In 1996, Taurel, then the chief operating officer being groomed to succeed CEO Randall Tobias, put together a program to slowly reduce the company's reliance on Prozac while pushing growth of other products. It was dubbed the Year X Plan, because no one knew exactly when it would be put to use. It wasn't clear at the time whether Prozac's patent would hold up until its expected 2003 expiration date.

Tobias put Taurel in charge of the Year X effort in part as a test. Since arriving at Lilly straight from graduate school in 1971, Taurel had distinguished himself as a deft manager. Throughout the 1970s and early '80s he was an ambassador of sorts, representing Lilly in marketing and managerial positions in Latin America, Eastern Europe, and France. In 1986, Taurel left his position in London as vice president of European operations and moved to Lilly's Indianapolis headquarters to head up the company's international operations. Though Taurel had ascended to great heights at Lilly, he had earned a reputation as a micromanager. "There was a time when he would sit in a meeting and ask a question, the next question, and the next one. And then he'd want more information after the meeting," says Tobias. "Often micromanagers have trouble delegating. They hover around the details and can't see the bigger picture. That's the way people might have seen Sidney years ago."

By 1997, Taurel was focused on the big picture, ready to begin implementing Year X. The plan's centerpiece was an industry-leading increase in R&D spending, to 19% of revenues. (The industry average is roughly 15%.) The effort also called for reducing the number of experimental failures and shortening the time each drug spent in various testing stages. Within a year, Tobias felt Taurel had proved himself as CEO material. He had grown, says Tobias, but most important, he had shaken his micromanaging instincts. "He learned to step back," says Tobias.

Taurel stepped back, but not too far. He became CEO in 1998. By 2001 and 2002, the company was enduring real hardships. Taurel was brutally frank about the adjustments that would have to take place. To communicate those changes, Taurel began broadcasting what he describes as his own "Larry King Live call-in show" on Lilly's corporate-wide television network. Twice a quarter he would sit before the camera and answer questions from Lilly's 43,000 employees. "It was important for everyone to see my optimism," he says. The show's main message was 'Don't cry over spilt milk,' " Taurel explains. "I think it empowers people when you are transparent, telling people where we're going and why."

The news wasn't easy to deliver. Taurel's increase in R&D spending necessitated a raft of austerity measures. In 2001, as the company's profits fell 7%, Taurel eliminated cash and equity bonuses for two years. In 2002 the company saw virtually no income growth. That year Taurel froze salary increases and even cut his own pay to $1. At the same time he struggled to temper Wall Street's short-term expectations while assuring investors that Lilly's labs would shine again.

Sure enough, they did. Taurel's spending boost helped Lilly fund deeper studies of compounds before they began clinical research. The tests helped Lilly increase the success rate of chemical compounds entering Phase I--the initial stage of clinical testing--from 50% to 90%. Last year Lilly launched three new drugs: Strattera for attention deficit and hyperactivity disorder, Forteo for osteoporosis, and Cialis, an impotence remedy that has already stolen roughly 20% of Viagra's market share and created a panic in Pfizer's marketing unit. In 2004, Lilly is expected to launch four medicines: Symbyax for bipolar depression; Alimta, a lung cancer medicine; Yentreve, for urinary incontinence; and Cymbalta, a depression treatment.

Lilly's fruitful labs are a scientific and financial anomaly these days, considering that the company is an old-school holdover from the days before its Big Pharma peers merged and morphed into lumbering behemoths. Competitors like Pfizer and GlaxoSmithKline--both more than three times Lilly's size in sales--are not nearly as productive as Lilly. Some observers insist that Lilly's recent fertile period is simply a matter of cyclical good fortune. This notion that it was Lilly's turn to win a round of R&D roulette angers Lilly's head of research and development, Steven Paul. "No, it's not luck," insists Paul. To illustrate what he has done as vice president of science and technology, he scribbles a cryptic equation onto a notepad. "We reduced the cost of getting a compound to Phase I and tripled the number of compounds, while adding more money to drug discovery by hiring new scientists." Exactly how Lilly accomplished that, Paul isn't telling. "The company that licks this arithmetic problem will be the leader in this industry," he says.

Taurel talks about the past few years with an odd fondness, the way some people recall having lived through the Great Depression. "While it was a trying time, it was also a time of great excitement," he says. "The problem with most companies is complacency. But there's nothing like a great challenge to motivate people." Taurel also takes credit for helping Lilly avoid the two biggest mistakes of competing companies. "We have not been distracted by mergers, and we didn't overspend on genomics research technologies," he says. "Some companies have invested too much and have not yet seen the full fruits."

Taurel's investments paid off, but the very thing that helped fund Lilly's R&D renaissance--bare-knuckle sales and marketing--may soon become a source of great pain for the company. During its drive for profits in the post-Prozac transition years, Lilly may have pushed the envelope in marketing some of its drugs, according to federal investigators. In March, the U.S. Attorney for the Eastern District of Pennsylvania opened a civil investigation into Lilly's marketing and promotional practices. The Justice Department charges that Lilly sales representatives improperly marketed Evista, an osteoporosis drug, for unapproved use as a breast-cancer preventive. It is also looking into whether the company marketed Prozac and Zyprexa (a bipolar maintenance drug) for unapproved uses. Lilly has issued a statement saying it is cooperating in the investigation and "cannot predict or determine the outcome of this matter."

The timing couldn't be worse. During the next year Lilly's marketing executives will be expected to turn over documentation and offer themselves up for questioning while simultaneously pulling off four product launches. Lilly executives won't discuss the investigation. Taurel does allow, however, that "we need more focus on being a good sales and marketing organization and to be reliable and trustworthy at a time when our industry is under such scrutiny." Taurel's biggest challenge in the months ahead will be changing the culture among his sales and marketing staff and correcting some of the problems that still exist in its production units. As for Lilly's factories, says Taurel, "in retrospect, we increased R&D at the expense of manufacturing."

Lilly's ability to fund the marketing efforts and the manufacturing overhaul may be threatened too. The Year X scheme relied heavily on what Taurel thought would be brisk sales growth of Lilly's biggest-selling existing drug, Zyprexa. But the $4.2-billion-a-year schizophrenia medicine is facing intense market competition and legal patent challenges. Zyprexa represents a third of the company's revenue. Strong competition from four bipolar-maintenance medicines is slowing Zyprexa's sales growth. After posting 19.5% growth in 2002, Zyprexa grew only 15.9% in 2003.

During the dark days of 2001 and 2002, Taurel put his faith in Lilly's neuroscience department. The hope was that it would fashion the next Prozac. And it did. Cymbalta, the crowning achievement of Lilly's recent drugmaking, breaks from the traditional so-called SSRIs by regulating levels not only of the neurotransmitter serotonin but of norepinephrine as well. Lilly is preparing to roll out Cymbalta this summer, but recent developments may prevent doctors from prescribing Cymbalta as broadly as Lilly had hoped. On Feb. 7, Traci Jones, a 19-year-old college student, hanged herself in a Lilly observation facility while involved in Cymbalta's human-testing phase. Last winter the FDA asked Lilly to conduct routine experiments on the effects on patients of higher-than-normal doses of Cymbalta. As part of that study, Jones had been given Cymbalta and then weaned off the drug for five days before her death.

The incident coincided with a growing concern about a link between antidepressants and suicide among children and teens. Last month New York Attorney General Eliot Spitzer opened an investigation of GlaxoSmithKline, charging that the company may have withheld data from clinical tests that show the dangers of antidepressant use by children (see box).

Lilly is cooperating with the FDA in an investigation into the Cymbalta clinical tests. But the incident is likely to play into the hands of FDA advisors, who last January urged the agency to warn doctors that some antidepressants may be linked to an increase in suicidal tendencies among youngsters. The FDA is reviewing data to decide whether the U.S. should join British authorities who in December recommended that doctors not prescribe certain SSRIs to children. Even so, Taurel believes the drug won't encounter any problems with the FDA and expects to get the agency's approval this summer.

Indeed, having pulled through the toughest part of Lilly's own Great Depression, Taurel finds there isn't much that dampens his spirits. He thinks R&D chief Steven Paul really has found a magic formula for more efficient drug development. And none of Lilly's existing drugs are scheduled to go off-patent until the middle of the next decade. "I'm an optimist," Taurel says. "I've always portrayed my confidence in the company." For Taurel, that part of the job is getting easier every day.

FEEDBACK jsimons@fortunemail.com