CHICAGO — When the coronavirus began sweeping around the globe this spring, people from Seattle to Rome to London canceled weddings and vacations, cut off visits with grandparents and hunkered down in their homes for what they thought would be a brief but essential period of isolation.

But summer did not extinguish the virus. And with fall has come another dangerous, uncontrolled surge of infections that in parts of the world is the worst of the pandemic so far.

A $5 rapid test for the coronavirus may be nearly as effective as the slower, more complex polymerase chain reaction test for identifying people who may spread the coronavirus, a novel experiment has found.

The study, conducted by scientists at the University of California, San Francisco, is among the first head-to-head comparisons of a rapid test and the P.C.R. diagnostic tool under real-world conditions.

But the number of participants was comparatively small, and the data have not been peer-reviewed or published. A rapid test still cannot conclusively determine that an individual is not infected; the tests are intended primarily to detect the presence of high levels of the virus, rather than its absence, and are authorized only to evaluate symptomatic people.

At the moment, anyone who has been exposed to the virus should be tested by P.C.R., said Joseph DeRisi, an infectious disease expert at U.C.S.F. and a co-leader of the project.