or the family of the healthy 19-year-old Bible student who hanged herself this month during a test of Cymbalta, an antidepressant being developed by Eli Lilly, the company's quick assurances that the death would not delay the drug's expected introduction this summer were wounding.

"Their posturing was offensive," said the Rev. Joel Barnaby, spokesman for Traci Johnson's family in Bensalem, Pa. "They were saying that this little death certainly should not defame our study or, God forbid, make us miss a deadline for F.D.A. approval."

While Lilly's reassurance might have struck the Johnsons as insensitive, the company has good reason to be concerned about deadlines.

Lilly has some of the most productive scientists in the drug industry, but also a disconcerting habit of delivering their discoveries later than investors expect.

In 2001, the company said that new drugs and the success of its popular antipsychotic medication, Zyprexa, would more than compensate for the loss of the patent on Prozac, its huge-selling antidepressant. But Lilly lost its hold over Prozac two years earlier than expected, and profit and revenue fell.

In 2002, the company said that soaring sales of a new drug to combat blood infections, as well as introductions of a Viagra-like pill and another for attention deficit disorder, would accelerate sales. But sales of Xigris, the blood infection medicine, have been disappointing, reaching just $160 million last year. And introductions of both Cialis, the impotence drug, and Strattera, the attention deficit pill, were delayed until 2003, in part because of manufacturing problems.

Last year, the company told investors that the introduction of Cymbalta would usher in an era of stellar growth. But Cymbalta's introduction was first delayed until the middle of this year and, with Ms. Johnson's death, may be pushed back again.

Suicides have occurred during trials of antidepressants. But Ms. Johnson was a healthy volunteer with no outward signs of depression or sickness. She hanged herself in a bathroom in Lilly's dormlike laboratory at Indiana University Medical School in Indianapolis.

"There is absolutely positively no observation, no knowledge, not even the suspicion that Traci was in a depressed state of mind or had a spirit wounded so much that she gave up on life," Mr. Barnaby said.

Rob Smith, a spokesman for Lilly, confirmed that Ms. Johnson was not depressed.

"We want to be very sympathetic to the family as they go through this tragic time, and our hearts go out to them," Mr. Smith said.

He added, "We're letting the facts guide us, but based on what we know at this point we don't think there'll be a delay to Cymbalta."

Still, the death of Ms. Johnson adds to worries among some researchers and physicians that antidepressants may cause a small number of people, particularly teenagers and children, to think about harming themselves. Just five days before Ms. Johnson's suicide, a federal advisory panel recommended that antidepressants carry warning labels describing the risk of suicide.

The result of all these delays is that Lilly's profits have been flat for years. In 2003, the company earned $2.37 a share, 4 percent lower than the $2.46 that it earned in 1999. And if Cymbalta is delayed beyond the summer, this year could be disappointing. In January, the company predicted that earnings for the year would grow as much as 10 percent.

While most analysts had hoped for more, they gush about Lilly's ability - all but unique, these days, among big drug companies - to discover new medicines. For now, many analysts believe that the company's share price is quite high, and some - doubtful about Lilly's ability to get all of its drugs to market as expected - say they are waiting for a better moment to buy.

"We want to own the stock, but the price is too high at the moment, and the momentum story isn't here yet," said Richard Evans, an analyst with Bernstein Research.

Mr. Smith said that no other drug company had survived the loss of its flagship medicine as well as Lilly. In 2000, Prozac sales were $2.2 billion, about 20 percent of the company's total. Within a year of Prozac's patent loss, the company surrendered 90 percent of those sales to generics.

"And while we have had delays in our pipeline, mostly because of manufacturing problems, we have made great strides to correct that," Mr. Smith said.

Yet Zyprexa, Lilly's mainstay since the loss of Prozac, is starting to show its age. In the last year, doctors have learned that its side effects include diabetes, stroke and death. Competing drugs with less serious side effects have been eating into Zyprexa's market share.

Some state Medicaid programs - by far the largest buyers of antipsychotics - have placed limits on the use of Zyprexa, the most expensive drug in the class.

Zyprexa brings in about a third of Lilly's sales and, many analysts estimate, half of its profits.

"Right now, this is the Zyprexa company, and I think Zyprexa has topped out," Mr. Evans said.

Zyprexa's patent is also being challenged in court, with a decision due sometime this summer. Most analysts expect Lilly to win the case, but a surprising - for Lilly, disastrous - outcome is still possible. Zyprexa's sales grew 16 percent last year, to $4.28 billion, largely on soaring sales outside the United States.

Sales of Evista, a hormone pill for osteoporosis, and Humalog, a branded version of insulin, may also start to erode. Evista, whose sales grew 12 percent last year, to $922 million, has been hurt by patients' concerns about hormone therapy.

Humalog has lost market share, but revenue nonetheless grew 22 percent last year, to $1.02 billion, as Lilly aggressively increased prices. Humalog competes with the Novo Nordisk drug Novolog. Four times in the last two years, both companies have raised prices on the same day and in identical amounts. Aventis is expected to introduce a competing insulin later this year, after which such lock step price increases may be harder to carry out.

All of which suggests that the drugs underpinning nearly two-thirds of Lilly's existing profit may have weak prospects for growth.

And what about its newest products? Strattera, Cialis and Forteo, an osteoporosis drug, had combined sales in 2003, their first year on the market, of $500 million, a healthy showing. This year, the company has introduced Alimta - for mesothelioma, a cancer of the lining of the lung - and Symbyax, a combination of Zyprexa and Prozac for bipolar depression. Lilly also hopes later this year to introduce a version of Cymbalta to battle stress-induced urinary incontinence, as well as a hospital version of Zyprexa. Next year, there is a promising diabetes medicine.

By far the most promising medicine in the company's pipeline is Cymbalta. Like Effexor, a pill sold by Wyeth that had $2.7 billion in sales last year, Cymbalta affects both serotonin and norepinephrine, neurotransmitters involved in depression. The dual action is thought to make Cymbalta especially effective against not only depression but pain as well.

Because its dosing regimen may be simpler, many analysts say that Cymbalta may outperform even Effexor. Sales could reach $4 billion by 2008, several estimate.

Cymbalta's promise is one reason that it is hard to bet against Lilly. Barbara Ryan, an analyst with Deutsche Bank Securities, switched her rating on the company to neutral after last year's delay of Cymbalta - only to see the company's shares rise.

"People like buying the stocks of drug companies that have new drugs," said Ms. Ryan, who still rates Lilly as neutral. "And Lilly has new drugs while most of the others don't."