For now, these researchers think this applies only to women who use or have used hormones such as birth control pills. And because there is so little other research on the question, the finding could turn out to be a statistical fluke.

Still, the idea is intriguing -- as is the suggestion from a handful of other studies that there may also be an optimal time in the cycle for a woman to have breast cancer surgery.

The hypothesis -- and it really is just that -- that a woman's cycle may affect diagnosis or treatment of breast cancer raises ``pretty interesting questions,'' says Dr. Jay Harris, chief of radiation oncology at Brigham and Women's Hospital and the Dana-Farber Cancer Institute.

And until recently, there have been few attempts at answers.

In fact, there's still no evidence that timing chemotherapy or radiation to a woman's cycle affects the outcome, says Dr. Ken Cowan, head of the medical breast cancer section at the National Cancer Institute. But there are three major studies now underway to see whether the odds of survival increase if a woman has breast cancer surgery at a particular point.

And there are already enough reasons, some specialists say, to recommend that women maximize their chances of an accurate mammogram by having it during the first half of their menstrual cycle, after a period is over.

``Although one never wants to take action on the basis of only one study,'' says Dr. Cornelia J. Baines, the University of Toronto epidemiologist who led the Canadian study, having a mammogram in the first half, or follicular phase, of the cycle ``can't do harm and may do good.''

Baines also believes that the apparent increased likelihood of missing a cancer in the second half of the cycle may partially explain why it has proved so hard to document a clear benefit of mammograms for women aged 40 to 49.

Nobody knows why mammograms would be harder to read during the second, or luteal, half of the cycle, she says, though it may be because of increased fluid and cellular activity in the breasts. In general, she notes, it's harder to read mammograms in women with denser breasts, but it is not clear whether breast density -- as opposed to swelling -- increases late in the cycle.

Dr. Dan Kopans, director of the breast imaging division at Massachusetts General Hospital, agrees it ``just makes good sense'' to have a mammogram early in the cycle.

The breasts are softer then, he says, which means there is less pain when they are compressed for the X-ray. And better compression yields more accurate mammograms.

However, many centers do not routinely schedule mammograms according to a woman's cycle, though some, like Newton-Wellesley Hospital, will do so if asked.

``It's very difficult to schedule screening mammograms with your cycle because they are booked in advance,'' adds Dr. Norman Sadowsky, director of the Faulkner-Sagoff Imaging and Diagnostic Center in Jamaica Plain. But it is ``very reasonable'' to try to schedule them in the first half of cycle because of the improved compression.

Research with a different type of test, magnetic resonance imaging, supports the idea that the second week of the cycle may be best for mammograms and breast self-exams -- and the fourth week the worst, he says.

A far dicier question is whether to schedule surgery to coincide with a presumed optimal time in the cycle. Even if proven desirable, it could be tricky to do because women with suspected cancer often have several surgeries over several weeks -- a biopsy to see if a lump is cancerous and later, removal of the lump or breast and some lymph nodes.

Still, the studies are intriguing -- and contradictory.

Nearly a decade ago, Dr. William Hrushesky, an oncologist at the Stratton V.A. Medical Center in Albany, N.Y., kicked off the debate with a study in mice that suggested the chances of a cure were doubled or tripled if breast tumors were removed around the time of ovulation.

In a follow-up 1989 study on 41 women, he found that survival was four times greater for women who had surgery at the time of ovulation and a week or so afterwards, than for those who had the operation closer to their periods.

No other study has confirmed that the time around ovulation is crucial, but other studies suggest there may be an advantage to surgery in the second half of the cycle. The theoretical reason for this -- and it's far from proven -- is that in the first half of the cycle, the hormone estrogen, which can drive some breast cancers, is ``unopposed'' in a woman's body. In the second half of the cycle, estrogen is balanced by the hormone, progesterone.

In early 1991, researchers at Guy's Hospital in London studied 249 women and found that the optimal time for surgery, at least for women whose cancers had spread to lymph nodes, may be in the second half of the cycle. The team went so far as to recommend scheduling surgery accordingly.

In September, 1991, Ruby Senie, an epidemiologist now at the Columbia University School of Public Health, also found in a study of 283 women that the best time may be the late luteal phase -- later in the cycle than Hrushesky had found.

Probably the strongest data in favor of the ``timing counts'' hypothesis comes from a 1994 Italian study of nearly 1,200 women by Dr. Umberto Veronesi. After eight years of follow-up, he found that in women whose cancer had spread to lymph nodes, those who had surgery in the second half of the cycle had a ``significantly better prognosis'' than those who had surgery in the first half of the cycle.

But other researchers find all this unconvincing, in part because the studies used different ways of figuring out where a woman is in her cycle. Some say it's enough to ask the woman to recall the date of her last period. Others believe it's necessary to measure hormone levels with blood tests or to do ultrasound exams of the uterus and ovaries to see if ovulation has occurred.

There are also studies that come to quite different conclusions, notably a 1994 Danish study of 1,635 women that found the timing of surgery had no effect on survival after 5 or 10 years.

Taken together, this mishmash means the provocative findings may be ``due to chance,'' says biostatistician Gary Clark of the University of Texas Health Science Center at San Antonio.

Dr. William Wood, chairman of the department of surgery at the Emory University School of Medicine, agrees. Some studies show it's best to have surgery in the first half of the cycle, he says, some that it's best in the second half, and there are ``three times as many studies showing it makes no difference.''

Yet the hypothesis won't go away.

In fact, it may even be gaining ground -- at least in the sense of being subjected to still more study, according to the Journal of the National Cancer Institute, which in April reported that three big, prospective studies are now in progress in Britain, Italy and the United States.

Will these studies settle the matter? Hrushesky thinks not, in part because the women will not be assigned surgical dates randomly. Instead, the researchers are assuming that women will come in for surgery at different points in their cycles.

Dr. Clive Grant, a surgeon at the Mayo Clinic who is a principal investigator for one of the studies, says it would be premature to randomize women to surgery at a particular time because the data are not yet compelling enough to warrant delay of surgery, as would inevitably happen in some cases.

So until better answers are in, what should a woman do?

Talk it over with your doctor, of course, and perhaps bear in mind the view of breast cancer guru Dr. Susan Love, adjunct professor at the UCLA School of Medicine. She says, through a spokeswoman, that she sees ``no harm from scheduling these patients [for surgery] during the early luteal phase.''

But should you do this if it means delaying surgery? ``That's exactly what we're trying to find out,'' says Grant.