The Centers for Disease Control and Prevention is embarking on an expansive study of the prevalence of novel coronavirus antibodies in people in 25 metropolitan areas, an effort to provide long-awaited insight into the way the virus is spreading and its presence in communities.

The study, which plans to test 325,000 people by fall 2021, will build on an antibody study that has been underway in six of those cities since March, according to Michael Busch, who is overseeing the study and is director of the Vitalant Research Institute. CDC spokeswoman Kristen Nordlund confirmed plans to announce the study but declined to discuss details.

Starting in June, this vast new antibody study will test samples from 1,000 blood donors each month for 12 months in the 25 metro areas. Researchers will test the samples for evidence of coronavirus antibodies, which are created by the immune system when someone is infected with the virus.

By determining who has antibodies, epidemiologists can figure out who has been infected with the virus, even if someone never reported a positive test or experienced symptoms. Several early studies suggest a large portion of those infected never display symptoms, making the extent of the virus’s spread more difficult to track.

“What we’re going to be able to do is a little like a monitoring station for forest fires. We’ll be behind the outbreak, but we’ll be able to tell policymakers whether opening up resulted in new spread,” said Busch, who added that case rates and death rates show a more limited, momentary picture of the infection’s spread.

A smaller, preliminary version of the study is underway, led by Busch and funded by the National Institute of Allergy and Infectious Diseases and the National Heart, Lung and Blood Institute. That study began testing 1,000 donors each month in six major metro areas: Boston, Los Angeles, Minneapolis, New York, San Francisco and Seattle. The CDC will provide $4.5 million in technical assistance as the study expands to 19 more metropolitan areas that remain to be finalized.

Vitalant Research is a nonprofit that specializes in blood-related research and is linked to 170 blood donation centers throughout the country. The effort will include major blood collection organizations such as the American Red Cross and New York Blood Center. Those groups will collect samples and demographic data from people who come to their facilities to make routine blood donations.

In each metro region, 1,000 blood samples will be tested each month for a year; then, six months later, 1,000 more blood donations from each region will be evaluated, allowing researchers to track the virus’s spread through the end of 2021.

Plans to launch the study were first reported by Reuters.

Antibody testing — known as seroprevalence surveys — can provide researchers and policymakers with a wider snapshot of the presence of the virus than the acute viral testing that has dominated much of the public health conversation for the last four months.

Instead of testing whether one person has the virus at a moment in time, which can often miss asymptomatic cases and gives no insight into prior infection, antibody testing determines whether someone ever had the virus — and, in some cases, when they had it. Prevalence and patterns of antibodies can sometimes indicate the difference between a more recent infection and one that occurred months ago.

Michael Mina, an assistant professor of epidemiology at Harvard University who is advising Massachusetts Gov. Charlie Baker (R) on that state’s coronavirus response, said using acute testing to inform public health policy is analogous to “trying to find a needle in a haystack.” Broad antibody testing, he said, “is the public health tool that should be used” because it provides evidence of whether the virus is spreading. If testing in a given month reveals 20 percent of New York City’s population has coronavirus antibodies, and testing a month later continues to show a prevalence of 20 percent, that suggests the city isn’t experiencing significant ongoing transmission.

“That probably means that the epidemic has diminished,” Mina said. “If it continues to increase, then you know that there’s still cases.”

The data can show how prevalent the virus is in a location, how incidence has changed over time and how it varies among demographic groups. In the case of the coronavirus, long speculated to exist asymptomatically in many of those infected, a survey such as this could shed light on the percentage of people infected with the virus who do not show symptoms of covid-19, the disease caused by the virus.

Some cities, including Boston and New York, have begun smaller-scale serological testing, in addition to the preliminary study being overseen by Busch. In Chelsea, Mass., a city adjacent to Boston that has emerged as a covid-19 hot spot, a survey of 200 blood samples found nearly a third tested positive for antibodies, according to reports. New York Gov. Andrew M. Cuomo (D) announced this month that 12.3 percent of samples taken from 15,000 New York state residents — including nearly 20 percent of samples taken from those living in New York City — were positive for antibodies.

Experts cautioned that because the planned nationwide study will rely on people who donate blood, it could limit the ability to provide a comprehensive picture of the virus’s spread.

“You’ll come out of [the study] knowing the fraction of people who have experienced this virus who are donating blood,” Mina said. “So that’s potentially a very limited pool.”

Busch said that because of the vast number of collection centers involved, donors are expected to cover a variety of demographic groups, which will allow researchers to extrapolate the data responsibly. He said the study will take into account that blood donors are generally healthier than the general population.

“What we know from many previous studies is that the data from asymptomatic blood donors will trail, by a month or two, the infection rates that are taking place in the general population,” Busch said. “They call it the healthy donor effect.”

Data collected from the tests will be made public on a rolling basis, allowing researchers and policymakers to track the changes in real time. Busch said the study could provide a sense of how close the country is moving toward herd immunity — a population’s resistance to the spread of a disease because a high proportion of individuals are immune to it — in the next year.

“If, after 15 months, we still see that less than 10 percent or so of the population has been infected, we’ll know we better have a vaccine because natural infection rates have not created herd immunity,” Busch said. “And that immunity won’t be enough to preclude another significant outbreak if society returns to the way we used to live.”