A steady decline in cases of toxic-shock syndrome recorded by the national Centers for Disease Control is prompting the Food and Drug Administration to reconsider whether to require a warning label on tampon boxes.

The agency proposed a warning label last year after disease experts investigating the newly discovered, sometimes fatal illness found that its victims almost always used tampons, particularly Procter and Gamble's Rely tampons. Rely was withdrawn from the market last September, and since then, at the agency's request, other manufacturers have included information about toxic shock on or in their packages.

Since Rely's withdrawal, each month has brought fewer and fewer case reports to the disease center. "In a way, we are having second thoughts because of the apparent decline in incidence," said an FDA spokesman.

But the center's investigators continue to emphasize that toxic-shock syndrome strikes users of all tampon brands, not just Rely, although no one knows how tampons promote the illness.

"People have misunderstood the business about Rely," said Dr. Arthur Reingold, an epidemiologist on the center's toxic-shock task force. "Many people are now falsely assuming that they are safe -- that they don't have to worry any more."

Reingold said the decline in recorded cases is probably due not only to the removal of Rely, but to the fact that fewer women are using tampons. He said when a nationwide outbreak of the disease was recognized a year ago, 70 percent of American women used tampons. By December, the proportion had falled to 55 percent. One manufacturer, acknowledging that there had been a decline in tampon use, said the market had leveled off and may be moving up again.

Some disease experts argue that the apparent drop in cases may be artificial -- suggesting that toxic-shock syndrome is no longer considered news and that fewer cases in the community are therefore being reported to health departments.

Among the skeptics are health officials in Minnesota, which has one of the most active disease surveillance systems of any state. There, toxic-shock syndrome has not fallen off. "We've not seen any change at all in the frequency of occurrence," said Dr. Michael T. Osterholm, chief of acute disease epidemiology at the state health department.

Osterholm said that in states where health department officials rely on local doctors, nurses and patients to call them, reports of toxic-shock cases have come in bursts each time the disease has received media attention. He said that because his department periodically calls health authorities around the state, his figure are less affected by publicity.

He said that in the first quarter of 1981, there were 17 or 18 cases in Minnesota. There were 17 in the last quarter of 1980 and 22 in the third quarter, during which the Rely publicity occurred.

Officials at Procter and Gamble, the makers of Rely, also contest the disease center's conclusion that toxic shock is on the wane. The company has been conducting its own telephone surveys of all 50 state health departments, gathering monthly totals that are ahead of the disease center's figures.

For January of this year, the departments reported 64 cases -- well above last year's January total of 43, according to Patrick Hayes, manager of the product section that includes Rely. He added that more cases occurred during the final quarter of 1980, after Rely's withdrawal, than during the spring quarter when it was on the market.

"Cases of toxic shock continue to be reported at high levels despite Rely's withdrawal," Hayes said.

No cases of toxic-shock syndrome have been reported in Virginia since February, according to Dr. Grayson Miller, the state epidemiologist. From November through February, one or two cases were reported each month.

The disease does not appear to have declined in Maryland. Six probable cases have occurred so far this year, compared with a total of 13 for 1980, according to Dr. Christina Lazar of the state's division of communicable diseases.

Two cases have been reported in the District of Columbia, one in December 1980 and one in January 1981, according to Dr. Philip Nieburg, an epidemiologist from the disease center assigned to the District. He said it was not clear why no cases were reported last summer, when the disease was at its height.

Nationally, the monthly total for toxic-shock cases peaked at 122 last August, according to Reingold of the federal disease center. The numers have fallen steadily in each succeeding month. For March of this year, the disease center recorded only 20 cases, and for April, it has reports so far of only two. As of May 15, the disease center had recorded a total of 1,128 cases of toxic-shock syndrome, 84 of them fatal.

"We are of the opinion that part of the decline is real -- that it's not all due to reporting," Reingold said. He said the constancy of the figures reported by Minnesota officials raised "the question of whether they are beating the bushes harder now than they were a year ago."

The FDA spokesman said the agency has reopened the comment period on its proposed warning label until June 29, not only because of the decline in cases, but also to consider recent discoveries about toxic-shock syndrome. Information that may be included in a revised warning includes Osterholm's finding that women who use high-absorbency tampons of any brand (usually labeled "super" or "super plus") have a higher risk of the illness than those using lower-absorbency tampons.

Osterhold said that since he reported that conclusion last January, some tampon manufacturers have made significant changes in their products' absorbency.

A spokesman for the Kimberly-Clark Corp., manufacturer of Kotex tampons, said his company plans to keep labeling its packages with cautionary statements "because the situation that caused us to put it on . . . really hasn't changed."

But he added that lack of uniformity among the current, voluntary warning labels may be confusing to shoppers. "There's a little bit of confusion among the brands," he said. "Some packages have it displayed right on the outside . . . and some don't have anything" except a package insert.

Thomas Moore, executive vice president of Tampax, Inc., said his company is inserting a statement on toxic-shock syndrome inside tampon boxes, and has updated its information to reflect new research findings. He said company surveys showed that the tampon market has dropped 20 percent since last summer, but that "it seems to have leveled off, and if anything is starting to increase . . . very slowly." He added that Tampax had not suffered a drop in sales.

The illness begins with fever, vomiting, diarrhea, and frequently headache and muscle aches. Within two or three days, a victim's blood pressure can drop dangerously low, producing dizziness or even shock -- blood pressure so low it causes organ damage.