Eli Lilly ( LLY)enters the first-quarter earnings season with a lot of promise, a big question mark and a desire to reverse three years of sliding earnings and meandering sales.

Perhaps that's why there are a lot of fence sitters on Wall Street: 15 analysts have hold ratings on the stock, according to Thomson First Call. There are nine buy recommendations and two sell recommendations.

And perhaps that's why analysts aren't straying very far from Lilly's financial guidance, offered in early January, when it said first-quarter earnings per share should be 65 cents to 67 cents. The consensus view is 66 cents with only a 1-cent variation on either side.

That EPS of 66 cents, or $715 million, would look dramatically good compared to the 38 cents, or $407 million, for the same period last year -- except for the fact that 2003's first quarter was clipped by 23 cents a share for restructuring costs and asset impairments.

Analysts predict first-quarter sales of $3.23 billion, up from $2.89 billion for the same period last year.

Wall Street is enamored with Lilly's research pipeline, which many analysts say is the most attractive among Big Pharma for near-term prospects. Their favorites include the depression drug Cymbalta, which they say may reach the market this summer; a diabetes drug, Exanatide, for which the company may seek Food and Drug Administration approval by midyear; and the stress urinary incontinence drug Duloxteine, which could be available in early 2005.

Another prospect is Alimta, already approved as a treatment for mesothelioma, a cancer of the lung lining. The company is seeking FDA approval for the drug to treat non-small cell lung cancer, which, analysts say, could significantly increase the drug's sales.

The company has a bunch of billion-dollar drugs -- Humalog and Humulin for diabetes and Gemzar for treating several cancers -- and it has gotten plenty of publicity from Cialis. That's the impotence drug that has grabbed second place in the U.S. market for new prescriptions after having been on sale only since late November. Lilly markets the drug with its creator Icos ( ICOS).

But analysts fret about a patent lawsuit affecting Lilly's anti-psychotic drug Zyprexa, which accounted for about 34% of the company's revenue last year and, by some Wall Street estimates, 40% or more of the earnings per share. Last year, Zyprexa recorded $4.28 billion in sales, up 16% from 2002.

Lilly, whose U.S. Zyprexa patent is good until 2011, is being challenged by Ivax ( IVX), Teva Pharmaceutical Industries ( TEVA) and Dr. Reddy Laboratories ( RDY).

As a result of all the pushing and pulling of sentiment, Lilly's stock has had a choppy 12-month existence, producing a total return -- assuming reinvestment of dividends -- of 20.8% for the 52 weeks ended April 14. That's more than double the 8.8% gain for the Amex Pharmaceutical Index, but well below the 29.7% gain of the S&P 500.

One analyst who got off the fence recently is David R. Risinger of Merrill Lynch. He considers the stock highly volatile but raised his rating March 31 to buy from neutral. "Eli Lilly has by far the best sales growth prospects among U.S. major pharmaceuticals over the next five years -- 11% vs. a group average of 5%," he said in a recent report to clients.

What prevents Risinger and other Lilly analysts from venturing into more rapturous comments is the Zyprexa patent challenge. Analysts and their legal consultants have spent considerable time poring over testimony in a trial that began in late January and ended in mid-February. But the parties still have a few months to file post-trial documents for the case that was heard in a federal district court in Indianapolis, Lilly's home town.

Risinger said a judge's decision is expected in the third quarter, "but a judgment could come at any time," he said. Any decision will be appealed, so the case "is likely to drag on for some time."

Risinger said he doesn't believe the generic companies have made their case against Lilly. Still, he offers dramatically different scenarios depending on the verdict. If Lilly wins, its 2005 earnings per share rises to $3.11 from his 2004 prediction of $2.85, and his price target climbs to $79.44 in 12 months. If Lilly loses, the 2005 EPS dissolves to $1.81 and the stock sinks to $42.96. (He doesn't own shares; his firm is a market maker in Lilly's stock and has had an investment banking relationship in the last 12 months.)

One analyst who jumped onto the fence in a favorable way is George Grofik, who recently upgraded his rating to hold from sell. Grofik now considers Lilly a medium-risk stock instead of a high-risk stock.

He told clients March 31 that he raised his rating because the stock was now trading in line with his price target and that Lilly has "the best pipeline in our coverage universe." In addition, his firm's legal consultants, after having reviewed trial testimony and certain depositions in the Zyprexa patent proceedings, have provided "increased comfort" that Lilly has a good chance of winning. If Lilly loses, he added, the stock could get whacked by 30%. (He doesn't own shares; his firm is a market maker in Lilly's stock.)

Another concern is that Zyprexa is showing some signs of slowing down. Oppenheimer analyst Scott R. Henry recently held his neutral rating on the stock but cut his EPS target to $2.78 from $2.85, saying that audited prescription data shows the Zyprexa franchise is weakening.

Henry said the company's assertions that it can provoke stronger U.S. growth for this franchise -- Zyprexa plus Symbyax, a drug that combines Zyprexa and Lilly's anti-depressant Prozac -- "does not appear reasonable given year-over-year prescription declines."

He made modest revisions in his U.S. Zyprexa/Symbyax sales predictions for 2004, 2005 and 2006. That led him to knock off 7 cents from this year's EPS estimate to $2.78, cut 10 cents from next year's estimate to $3.33, and trim 12 cents from 2006's EPS prediction to $4.10. "The Zyprexa overhang mutes the new drug launches," Henry added. (He doesn't own shares; his firm doesn't have an investment banking relationship.)

Lilly got a boost March 30 from the FDA, which approved an injectable version of Zyprexa. Ken Kulju, of Credit Suisse First Boston, told clients recently that FDA action had been expected in late 2002 but was delayed due to the agency's inspection of the company's manufacturing operations. (He doesn't own shares; his firm has had an investment banking relationship with Lilly in the last 12 months.)

Although the injectable Zyprexa will help Lilly "defend its current position," Kulju cautioned clients about recent unfavorable prescription data. "We believe Zyprexa's higher incidence of weight gain has been one of the primary contributors to the recent market share losses to newer antipsychotic agents," said Kulju, who has a neutral rating on the shares.