It's now or never for physicists' favorite explanation of dark matter, the invisible material that seems to pervade the universe. The largest, most sensitive search yet for the particles many physicists think make up dark matter—weakly interacting massive particles (WIMPs)—will begin in March at the XENON1T experiment at the Gran Sasso National Laboratory in Italy.

The project is the latest in a line of detectors that date back as early as the 1980s but have all failed so far to find dark matter. If the elusive particles go unfound in the next few years at XENON1T, physicists may have to abandon the leading theory and search for more exotic explanations. “Our best models are within reach of XENON1T,” says Rafael Lang, a physicist at Purdue University who works on the experiment. “If we don't see it, that means our ideas are completely wrong, and we really have to go back to the drawing board.”

WIMPs are a prediction of superstring theory. This extension of the Standard Model of particle physics proposes the existence of partner particles for all the known fundamental bits of matter in the universe. WIMPs would be the lightest of these partners, and physicists favor them because the theory naturally predicts just about the amount of dark matter that experts know must exist because of its gravitational pull. (Dark matter represents an estimated 84 percent of all matter in the cosmos.) Many versions of WIMPs have already been ruled out because previous searches for them turned up nothing, but investigators are still hopeful that one of the remaining possibilities will show up.

Buried in a cave 1,400 meters underground, XENON1T houses a large cylindrical vat filled with 3,500 kilograms of liquid xenon. The substance naturally gives off light when its atoms are disturbed; scientists are aiming to catch the rare occasion when a dark matter particle collides with a xenon nucleus, an impact that should leave a unique energy signal. Although dark matter is thought to be ubiquitous—roughly 100,000 dark particles fly through every square centimeter of space each second—it almost never interacts with regular matter and generally makes its presence known only through gravity. After the planned two-year search at XENON1T, the detection of just 10 particles that appear to match dark matter's predicted properties would be enough to claim a discovery.

The $15-million project, sponsored by a collaboration among 10 different countries, follows a previous iteration of the experiment that was 25 times smaller. The new XENON's larger collecting volume, as well as improved shielding to block other particles that might masquerade as dark matter, should allow it to surpass the earlier experiment's level of sensitivity within two days of turning on. It should also overtake the current leading dark matter experiment, the 370-kilogram Large Underground Xenon experiment (LUX) in South Dakota, within weeks. “I would not at all be surprised if XENON1T were able to make a discovery that had just barely escaped the generations of experiments that came before it,” says Tim Tait, a theorist at the University of California, Irvine, who is not involved in the experiments.

Meanwhile WIMPs could also show up any day now at the Large Hadron Collider near Geneva, where protons crash into one another at near the speed of light to give rise to new particles. The accelerator began a second run last year at almost twice the energy with which it turned on in 2009 and now should be powerful enough to create roughly the same range of WIMPs that might be detectable at XENON1T.

And if in the next few years, neither of them sees a hint of the particles, the time may come for theorists to move on to another explanation for dark matter. “On one hand, we know it exists, but on the other hand, we know very little about it, so it's very easy to theorize about possibilities,” Tait says. “If we don't see it, that tells us the dark matter has turned out to be more weird and wonderful than we had originally guessed it might be.”