Physics has missed a long-scheduled appointment with its future—again. The latest, most sensitive searches for the particles thought to make up dark matter—the invisible stuff that may compose 85 percent of the mass in the cosmos—have found nothing. These elusive particles, called WIMPs (weakly interacting massive particles), may simply be better at hiding than physicists thought. Alternatively, they may not exist, which would mean that something is woefully amiss in the underpinnings of how we try to make sense of the universe. Many scientists still hold out hope that upgraded versions of the experiments looking for WIMPs will find them, but others are taking a second look at conceptions of dark matter long deemed unlikely.

The first null result this summer came from the Large Underground Xenon (LUX) experiment, a third of a metric ton of liquid xenon held at a frosty −100 degrees Celsius inside a giant, water-filled tank buried one and a half kilometers under the Black Hills of South Dakota. There, shielded from most sources of contaminating radiation, researchers have spent more than a year's worth of time looking for flashes of light emanating from WIMPs striking xenon nuclei. On July 21 they announced they had seen none.

The second disappointing report came on August 5 from the most powerful particle accelerator ever built: CERN's Large Hadron Collider (LHC) near Geneva. Since the spring of 2015, the LHC has been pursuing WIMPs by smashing protons together at unprecedentedly high energies, at rates of up to a billion collisions per second, pushing into new frontiers of particle physics. Early on, two teams had spied a telltale anomaly in the subatomic wreckage: an excess of energy from proton collisions that hinted at new physics perhaps produced by WIMPs (or, to be fair, many additional exotic possibilities). Instead, as the LHC smashed more protons and collected more data, the anomaly fizzled out, indicating it had been a statistical fluke.

Taken together, these two null results are a double-edged sword for dark matter. On one hand, their new constraints on the plausible masses and interactions of WIMPs are priming plans for next-generation detectors that could offer better chances of success. On the other, they have ruled out some of the simplest and most cherished WIMP models, raising fresh fears that WIMPs might be a multidecadal detour in the search for dark matter.

Edward “Rocky” Kolb, a cosmologist now at the University of Chicago who in the 1970s helped to lay the foundation for WIMP hunts, has declared the 2010s “the Decade of the WIMP” but now admits the search hasn't gone as planned. “We are now more in the dark about dark matter than we were five years ago,” he says. So far, Kolb notes, most theorists have responded by “letting a thousand WIMPs bloom,” creating ever more baroque and exotic theories to explain how the supposedly ubiquitous particles have dodged all our detectors.

Theorists have two intertwined reasons to hunt for WIMPs. The first is that WIMPs are a natural consequence of the most popular extensions to the Standard Model of particle physics, which predicts their production shortly after the big bang. The second is that if such primordial WIMPs exist, straightforward calculations suggest their present-day abundance and behavior should now almost exactly match the quantities and qualities of dark matter inferred from observations. This so-called WIMP miracle has sustained the search for decades, but now some theorists are questioning its validity.

For example, in 2008 Jonathan Feng and Jason Kumar, both then at the University of California, Irvine, showed how a phenomenon known as supersymmetry could produce a hypothetical class of particles much lighter and more weakly interacting than WIMPs. “These particles result in the same amount of dark matter we see today, but they aren't WIMPs,” Feng says. “This upsets the apple cart because it is just as well motivated theoretically. We call it the WIMPless miracle.”

The decaying theoretical underpinnings for simple WIMP models, paired with the growing list of empty-handed detection efforts, have led Feng and many others to propose that WIMPs are part of a more complicated picture: a hidden realm of the universe filled with varieties of dark particles interacting with one another through a suite of dark forces, perhaps exchanging dark charges through bursts of dark light. Because they offer theorists many more variables to play with, such “dark sector” models can be reconciled to fit into the ever tighter straitjacket of facts placed on dark matter by new data—but the downside is that this sprawling flexibility makes them very difficult to conclusively test.

“With the dark sector, you're free to invent almost whatever you want,” says David Spergel, an astrophysicist at Princeton University. “Now that we have lost the guidance from the WIMP miracle, the space of available models is huge. It's a playground where we don't know what the right choices are—we now need more hints from nature about where to go next.”

Some physicists, following nature's hints, have abandoned WIMPs altogether. For instance, ghostly particles called neutrinos are known to exist and come in three varieties, or flavors. Although the three varieties are not massive enough to account for dark matter, by virtue of having mass at all they open the possibility for the existence of a fourth—a massive, so-called sterile neutrino. “Almost all neutrino mass–generation mechanisms require the existence of sterile neutrinos, and it would be very easy for some of these sterile neutrinos to account for the dark matter,” says Kevork Abazajian, a theorist at U.C. Irvine.

Another perennial dark horse candidate for dark matter is the axion, a hypothetical weakly interacting particle first postulated in 1977 to explain and resolve otherwise mysterious asymmetries in quantum interactions. For axions to explain dark matter, they would need to occupy a relatively narrow range of masses and be far lighter than WIMPs. “If we don't find the WIMP, theorists will just switch their bets to axions,” says Peter Graham, a physicist at Stanford University.

Beyond WIMPs and dark sectors, sterile neutrinos and axions, there are even more exotic possibilities for dark matter, although they occupy the fringes of physics, including “primordial” black holes, extra dimensions and the possibility that Einstein's theory of gravity is wrong in some way.

Whatever their preferred candidate might be, the biggest concern for many physicists grappling with dark matter is not that the concept will eventually be seen as somehow invalid or entirely mistaken—the observational evidence for dark matter's existence is overwhelming. Instead they worry that dark matter's identity might simply prove to be irrelevant to other great mysteries in physics and thus offer no new paths toward understanding the true nature of reality.

“The desire is for dark matter not only to exist but also to solve other outstanding problems of the Standard Model,” says Jesse Thaler, a physicist at the Massachusetts Institute of Technology. “Not every new discovery can be a revelation … where afterward theories suddenly fit together much better. Sometimes new particles just make you say, ‘Who ordered that?' Do we live in a universe where each discovery leads to deeper, more fundamental insights, or do we live in one where some parts have rhyme and reason, but others don't? Dark matter offers either possibility.”