n the late 1980's, a blood test for prostate cancer came into widespread use. The test, the P.S.A., can find cancers when they are too small to feel and before they have spread. It is only now being evaluated in clinical trials to see if it reduces the risk of dying from the disease. Gina Kolata asked two experts to assess national data on prostate cancer.

DR. IAN M. THOMPSON JR., a urologist and professor of surgery at the University of Texas Health Science Center at San Antonio, looks at the data on P.S.A. screening and is encouraged.

Before the P.S.A. test came into use, the cancer's incidence was steady from year to year. That changed with increasing use of the P.S.A., for prostate specific antigen. The number of prostate cancer cases diagnosed rose from 143.3 per 100,0000 in 1990 to a peak of 195.6 per 100,000 in 1992 and then fell to 155.3 per 100,000 in 1998, the most recent year for which data are available.

The pattern reflects the use of the test, which found men who had cancer and had never before had any kind of test sensitive enough to find it. The incidence fell as more and more men who had had cancer all along were tested and their cancers found.

Before the blood test was introduced, Dr. Thompson said, doctors looked for the cancer by rectal exams, which he said was "not very good." The increasing incidence shows, he said, that the P.S.A. is doing what it is supposed to do — finding cancers early.

As for the mortality rates, Dr. Thompson said they were falling steadily and dropped by 16 percent from 1990 to 1998, going from 38.6 per 100,000 to 32.3 per 100,000. Treatment improved in that time, but Dr. Thompson said that alone was not enough to explain the drop in the death rate. Prostate cancers that have spread to other organs are less common, dropping in incidence by 60 percent from 1990 to 1998.

J. Michael Short for The New York Times Dr. Ian Thompson Jr. says P.S.A. tests are finding prostate cancers early.     Interactive Graphic: Weighing the Value of Prostate Antigen Tests   Topics  Alerts Cancer Medicine and Health Create Your Own | Manage Alerts Take a Tour Sign Up for Newsletters Join a Discussion on Men's Health     Marty Katz for The New York Times Dr. Otis Brawley says screening for prostate cancer may not save lives.

"Metastatic disease is falling so low that we've never seen those numbers before," he said. That, he says, is a tribute to early detection with the P.S.A. test.

"Yes, there are probably some people who are treated unnecessarily," he said. "But all of medicine is that way. Is there a chance that when you have your appendix removed it will turn out to be normal? Yes."

Dr. Thompson is 47 and has had a P.S.A. test; it was normal. He intends to continue having the test and to be treated if cancer is found. The treatment, he said, does not worry him. It is a small price to pay for a potential cure.

DR. OTIS BRAWLEY, professor of medical oncology and epidemiology at the Winship Cancer Institute of Emory University, is concerned. National data, he said, make him wonder whether the P.S.A. test is saving many lives.

The test, Dr. Brawley said, has led to a huge increase in the number of prostate cancer diagnoses.

But if the P.S.A. was saving lives, the early diagnosis of prostate cancer should lead to a sharp downturn in its death rate. That has not happened, Dr. Brawley said. While the mortality rate has fallen, it did not fall precipitously and the introduction of better treatments clearly played a role.

Dr. Brawley said the data indicated that the test was finding tumors that would not be noticed and would cause no medical problems. But these tumors look exactly like dangerous ones.

"There are cancers that fulfill all the histological criteria for malignancy," Dr. Brawley said. "When a pathologist looks under the microscope, he says, `Yes, this is cancer.' " But these tumors are not dangerous, and many that are dangerous may not be cured by early diagnosis and treatment.

"The concept that every cancer that can be found early can be cured is a faulty concept," he said.

Yet, with no way of knowing which tumors are potentially lethal, doctors usually treat everyone with methods like surgery or radiation that can cause impotence and incontinence.

While the number of men with metastatic disease at the time of diagnosis has gone down, Dr. Brawley says the reason is not necessarily that the P.S.A. test is finding cancers when they can be cured.

Doctors, he said, operate and remove the prostate only when they find no evidence that cancer has spread. Those men are counted as having localized disease. But, he said, within two years, 40 percent of them have recurrences. The cancer had already metastasized, but the spread was not obvious when the original diagnosis was given. Before the test, most of those cancers would have been found after they had spread and counted as metastatic.

Dr. Brawley worries that some doctors are including P.S.A.'s in routine blood tests without discussing these issues.

"If a man understands that the test is unproven and still wants it, he should be allowed to get it and should not be criticized," Dr. Brawley said.

At 42, Dr. Brawley has not been tested and does not want to be. Once, a doctor ordered the test on a lab order sheet without telling him. "I asked her to scratch it off," he said.