LOS ANGELES — In deep underground laboratories around the globe, a high-tech race is on to spot dark matter, the invisible cosmic glue that’s believed to keep galaxies from spinning apart.

Whoever discovers the nature of dark matter would solve one of modern science’s greatest mysteries and be a shoo-in for the Nobel Prize. Yet it’s more than just a brainy exercise. Deciphering dark matter — along with a better understanding of another mysterious force called dark energy — could help provide insight into the fate of the universe.

Previous hunts for the hypothetical matter have turned up nothing, but that has not deterred some two dozen research teams from plumbing the darkness of idled mines and tunnel shafts for a fleeting glimpse.

Dark-matter detecting machines today are more powerful than previous generations, but even the best has failed so far to catch a whiff of the stuff. Many teams are now building bigger detectors or toying with novel technologies to aid in the hunt.

“We’re in the golden age of dark matter search,” said Sean Carroll, a California Institute of Technology theoretical physicist who has no role in the experiments. “It’s looking good for some breakthroughs to happen.”

Scientists acknowledge they are still in the dark about dark matter. The prevailing theory is that it’s made up of tiny, exotic particles left over from the Big Bang some 13.7 billion years ago. Dark matter, thought to make up a quarter of the universe’s mass, gets its name because it doesn’t give off light or heat. Astronomers know it exists because of its gravitational tug-of-war with stars and galaxies.

Knowing that dark matter exists is a far cry from knowing what it is. Most experiments are searching for theoretical particles called WIMPS — or weakly interacting massive particles — the leading dark-matter candidate.

The underground custom-built machines are all waiting for the rare moment when a WIMP hits the atomic nucleus and causes an elastic recoil. Experiments have to run below ground to prevent cosmic rays from interfering with the results.

Dark matter researcher Neil Spooner of Sheffield University in England sums it up this way: “You have a needle in a haystack and you’re trying to remove the hay. You need better technology to pull out the event you’re looking for and reject the rubbish.”

Subterranean experiments are humming in an idled iron mine in Minnesota and in caverns in Canada, England, France, Italy, Japan, and Russia. Last month, the National Science Foundation chose the defunct Homestake gold mine in South Dakota to be the site of one of the largest and deepest labs of its kind in the world — bigger than six Empire State Buildings stacked below ground.

The competition is cutthroat, and physicists spar over which technology works best.

The front-runner for the past several years, called CDMS for cryogenic dark matter search, uses ultracold silicon and germanium crystals each the size of a hockey puck to sift out telltale vibrations of a WIMP collision. Newer contraptions use noble gas such as xenon or emerging technologies like superheated liquid bubble chambers.

“There’s no perfect dark matter experiment or detector. All of them have their quirks and limitations,” said Juan Collar, a particle physicist at the University of Chicago and part of a team called COUPP.

Scientists realize they may be in for a reality check.

“It’s possible that no matter how big of an experiment you build, you may not find anything,” said Steve Ahlen of Boston University, who along with collaborators from the Massachusetts Institute of Technology and Brandeis University, is building a prototype that will be placed underground next year in a yet to be determined location.

There have been false alarms. In 2000, Italian scientists working in an underground lab near Italy’s Gran Sasso mountain range claimed to have detected a dark matter signal. But no one has been able to reproduce the result, and the claim is not widely recognized in the scientific community. The Italian researchers have since been working on a second-generation detector and expect to present new results next year.

This spring, a rival group led by Columbia University’s Elena Aprile, who also works in Gran Sasso, shocked her peers by announcing at a science meeting that her liquid gas project called XENON10 is more sensitive and rejects more background noise than the CDMS detector.

“The more sensitivity you have, the closer you get to the truth,” Ms. Aprile said.

CDMS spokesman Bernard Sadoulet of the University of California, Berkeley, said it helps to have more than one technology searching for dark matter to cross-check results. He added that his team has been taking data with its scaled-up detector since last year and expects to regain the sensitivity lead.