Yes. Since progesterone is safe, and advocates say it has helped thousands of women, I think it's worth trying. As a reproductive endocrinologist, my first approach is to suppress ovulation. But the hormones used may have unacceptable side effects. By contrast, progesterone has minimal side effects. It may be the most benign active medication for PMS.

Scientific studies, especially controlled ones, haven't shown significant benefit. But many were riddled with flaws and used low doses: 200 to 400 milligrams vs. the 1,200 to 1,600 mg usually given for PMS.

Such trials don't prove typical doses are no better than placebo. Also, the hormone may help a subset of patients we haven't yet identified.

Progesterone has been shown to inhibit ovulation in some women. It has strong central nervous system effects, acting as a sedative or anesthetic. Besides being a diuretic, progesterone is thought to counteract some of estrogen's harmful effects. And it's metabolized into many other active substances.

Yet drug companies have no incentive to do clinical research on progesterone. Since it can't be patented, it's not worth spending millions to meet FDA requirements. And without FDA approval, doctors hesitate to try it.

Just because we don't know the mode of action is no reason to avoid a safe agent that many women favor.

Let's not repeat history. Women had to suffer menstrual cramps for too long. We didn't take them seriously until the prostaglandin mechanism was discovered. -- Dr. Wayne S. Maxson director, Northwest Center for Infertility and Reproductive Endocrinology, Margate, Fla.

No. There's no scientific evidence that progesterone is any better than a placebo for PMS.

It's a frustrating disorder that ob-gyns often refer to psychiatrists. No specialty has taken responsibility for it.

I believe there's more than one type of PMS. Progesterone might help a subgroup of patients. But we don't know enough yet to identify the subgroups or explain how the hormone might work.

Most important, we've found that very few of the patients who come to us complaining of severe premenstrual symptoms actually have cyclic PMS that meets strict diagnostic criteria.

We follow each patient for at least two months to observe whether recorded daily symptoms correspond with cyclic changes. You can't make the diagnosis retrospectively.

Patients come in weekly. They give us their symptom records and have a chance to talk with each other about their complaints. Many say that meeting women with similar symptoms helps.

There's no formal psychotherapy. But all are encouraged to make at least one healthful change -- following a better diet, aerobic exercising, quitting cigarettes or cutting down on alcohol, caffeine or stress -- that may increase their ability to function.

Our follow-up shows that 80 percent of the women don't have PMS. Their symptoms aren't cyclic and usually resolve just with the program's support. Most who do have cyclic symptoms also improve.

Only about 5 percent fail to respond and are considered for therapy with psychoactive drugs or progesterone. And it's our impression that the hormone's benefits don't last longer than six months. Patients seem to revert to their original state or worse. -- Dr. Leslie H. Gise associate professor of psychiatry and director of the Premenstrual Syndromes Program, Mount Sinai School of Medicine, New York City