The study's authors say it demonstrates that losses used to justify big premium hikes have been overstated. The Foundation for Taxpayer and Consumer Rights is funded in part by tort lawyers who sue doctors and hospitals in malpractice cases, as well as sue corporations in product liability cases.

"We're not saying they shouldn't use estimates, but it's how far off they are," said Harvey Rosenfield, a lawyer for the foundation. "It's out of the ballpark, not even close."

Insurance industry representatives dismiss the study's methodology, saying it selectively uses years when the industry overestimated its losses and ignores more recent years when the industry wound up paying more than it reported initially to regulators. Lately, they add, insurers' losses have outstripped estimates from only a few years ago.

"They cherry-picked the years," said Lawrence E. Smarr, president of the Physician Insurers Association of America.

Insurance representatives also say insurers rebated or lowered premiums during the years that losses were overstated. They note that at least half of the medical malpractice insurance market is written by doctor-owned mutual companies, which, they say, have no incentive to overcharge.

The study arises out of a long-running and polarized debate about malpractice losses and how to insure against them. On one side, insurers and doctors argue that the legal system is out of control and tighter restrictions are needed on how much plaintiffs can recover. On the other side, consumer groups and trial lawyers argue that the insurance industry manipulates prices and the medical profession doesn't do enough to control medical injuries and that those industries need closer supervision.

About 190 companies offer medical malpractice insurance, writing about $9.4 billion in policies a year. Traditional insurance accounts for only about a third of the market, according to some estimates, because many large providers rely on self-insurance and other arrangements. The United States spends about $1.7 trillion on health care annually.

Proponents of tort reform often present spiking medical malpractice premiums, which can be in the six figures for some specialties, as Exhibit A in arguments for capping damages, an argument that has resonated. So far, 25 states have passed measures that restrict legal remedies in tort cases in some fashion.

Earlier this year, for example, Maryland imposed a freeze on noneconomic damage awards of $650,000 for the next four years and reduced limits on noneconomic damages in wrongful death suits to $812,500 from $1.6 million. President Bush is urging Congress to pass a sweeping medical malpractice bill, approved by the House but defeated in the Senate last year, that would cap noneconomic damages and provide an alternative system of dispute resolution.

The debate hinges on the murky question of how insurance prices are set. In theory, rates are set by state regulators based on insurers' past losses, actuarial estimates for the future and a reasonable rate of return. In practice, prices vary most according to the number of insurers scrambling for business. When profits are high -- for example, when bond and stock markets are thriving and insurers are making money on investments -- premiums tend to fall as new competitors rush in and compete for premiums to have money to invest. When investment returns dwindle or big losses hit, insurers exit the market and prices rise, a point conceded even by insurance representatives.

But big jury awards certainly drive up underlying malpractice costs and contribute to higher premiums. Insurers and doctors say that the number of $1 million-plus claims as a percentage of all claims has doubled to 8 percent in the last five years. What's more, some jurisdictions -- Madison County, Ill., is often cited -- return much higher awards than others, making underwriting difficult. And both sides concede that malpractice claims are particularly tricky to estimate because of the long time it takes to resolve them and juries' unpredictability.

Neutral experts are rare -- and split in their opinions. Tom Baker, director of the Insurance Law Center at the University of Connecticut School of Law, argues in a new book, "The Medical Malpractice Myth," that the number of malpractice lawsuits is relatively low considering the medical injury problem and, as the new study argues, that incurred losses jump in response to insurers' competitive needs. But Randall R. Bovbjerg, of the Urban Institute, argues that the doctor-insurer side is "more correct" because underlying costs from lawsuits are in fact rising.

Fueling frustration on all sides: a surprising lack of data available from the nation's patchwork system of state regulation. Two years ago, the Government Accountability Office studied the problem and said that although insurance losses "appeared to be the greatest contributor" to higher premiums, a "lack of comprehensive data . . . prevented us from fully analyzing the composition and causes of those losses." The study concluded that Congress "may want to encourage" state regulators to collect data on "the frequency, severity and causes of losses on medical malpractice claims."

Even how much doctors actually pay in premiums is not readily available. Annual rates are kept by individual state insurance departments, and all sides rely on an industry newsletter, Medical Liability Monitor, based in Chicago, to collect data.

The new study was written principally by Rosenfield. A 53-year-old lawyer, consumer activist and protege of Ralph Nader, Rosenfield is a familiar and controversial figure in California political circles and was the author of Proposition 103, a sweeping series of insurance regulations that passed over bitter industry opposition in 1988.

Also contributing was Jay Angoff, a former Missouri insurance director and industry critic who wrote a study negative on the industry last summer that itself has drawn heavy criticism from, among others, the National Association of Insurance Commissioners, the regulators' Indianapolis-based trade group.

Rosenfield said the study shows that incurred losses can be manipulated to support insurers' need for higher prices when their stock and bond investments go bad. He said that the chronic overestimation came during years when the industry complained loudly of a tort crisis that did not actually exist and that regulators do not go back to check what happened to losses after approving rate hikes.

Smarr, of the physician-owned insurers' group, said actuarial estimates were off in the late 1980s and early 1990s because insurers didn't anticipate a drop-off in the frequency of claims that began after 1984. He said, though, that insurers paid more than they estimated on claims starting in 1998, because the cost of defending claims and paying damages has jumped. He said the activists' study selects only years that support their points.

"Juries are making larger and larger awards, and that drives settlements," Smarr said.

According to industry data, insurers estimated in 1998 that year's claims would cost $5.2 billion but have lately told regulators they have already actually paid $5.3 billion on 1998 claims and expect to pay a total of $6 billion when all are resolved.

Angoff, the former Missouri insurance director who contributed to the advocates' study, counters that insurers did not need to raise premiums during the mid to late-1990s because the booming stock market provided incentives to keep rates low to bring in more premiums.

What is certainly up in the air: the outcome of the current malpractice crisis, which has triggered new calls for tort reform and protests by doctors. In recent years, insurers have reported another big jump in incurred losses to $7.1 billion in 2004, up 43 percent from 2000. The figures since 2000 are still too new to know how much will actually be paid.

But premiums jumped again. According to the Medical Liability Monitor, the highest premium for a Miami general surgeon more than doubled from 2001 to 2004, to $277,000 a year.