The next time you look up at the night sky and find yourself marveling at the number of stars overhead, know that you are only seeing part of the magnificent bounty that our galaxy holds. Most of those Milky Way stars are not isolated orbs. Rather an average star has at least one planetary companion, invisible to the naked eye and in most cases as yet unseen by telescopes, according to a new analysis.

That extrasolar planets should be even more common than stars, which themselves seem innumerable, lends support to the hope that somewhere up in the night sky, circling one of those stars, is a world like Earth where life may have had a chance to take root, and maybe even have evolved into an intelligent form.

The analysis of planetary frequency in the Milky Way appeared in the January 12 issue of Nature. (Scientific American is part of Nature Publishing Group.) The researchers, led by astronomer Arnaud Cassan of the Paris Institute of Astrophysics at University Pierre and Marie Curie, used a small sample of planetary discoveries to infer the size of the overall planetary population. Extrapolating from a few known planets and the relatively low probability that each of those planets should be detectable from Earth, the researchers found that each star is home to an average of 1.6 planets.

The process is a bit like estimating the average number of children in a typical family by peering into a handful of random homes, counting the number of children in view, and estimating how many more are at school or otherwise out of sight. As such, the planetary demographics are still rudimentary; given the small-number of statistics, the actual average could be closer to one planet per star, or it could be well over two planets per star. But the general ubiquity of extrasolar planets, which other astronomical campaigns have also suggested in recent years, seems unassailable.

"This is not a surprise, but it's a really interesting thing to know," says astronomer Scott Gaudi of The Ohio State University, who did not contribute to the new research. Perhaps most encouraging is the finding by Cassan and his colleagues that the frequency of planets rises as the mass of those planets decreases. Large planets akin to Jupiter are relatively rare, midsize planets such as Neptune are present around roughly 50 percent of stars, and small planets just five to 10 times the mass of Earth are even more numerous than that. "Planets are common, and low-mass planets are as common as dirt in some sense," Gaudi says.

Cassan based the galactic census on a planet-finding method called gravitational microlensing. Using the Warsaw University Telescope in Chile, astronomers monitor roughly 200 million stars to look for the sudden and anomalous amplification in the light from any one of them. That brightening can be caused by another star passing in front of the background star, with the gravitational field of the intervening star acting like a lens to focus the light of the background star toward Earth. Such alignments are rare, but by monitoring so many stars for years on end, the campaign, known as the Optical Gravitational Lensing Experiment (OGLE), has recorded thousands of microlensing events.

The brightening and subsequent dimming of the background star due to microlensing does not always follow a smooth bell curve, however. In about a dozen cases identified by OGLE and by the similar Microlensing Observations in Astrophysics (MOA) experiment based at Mount John University Observatory in New Zealand, irregularities in the lensing signal point to a planet orbiting the foreground star and distorting the symmetry of the lens. The duration of a deviation from the bell curve indicates the suspected planet's mass.

Microlensing has its downside—the planetary signals are ephemeral, lasting only as long as the background star and the planet-hosting star remain in alignment (typically about a month). But it has one critical advantage over other planet-hunting techniques: it is sensitive to bodies not especially close to their parent stars. More prolific planet-search methods, including the technique employed by NASA's Kepler spacecraft, which detects periodic variations in starlight caused by orbiting planets eclipsing their stars, have the most success detecting planets that orbit very close to their host stars and hence complete an orbit very quickly.

"Microlensing can probe planets of all masses for a very large range of orbital separations," from about 0.5 times to 10 times the Earth–sun distance, Cassan says. He notes that the abundance estimates can only increase with exploration of a larger range of orbital distances and planetary masses. "Our results are given for masses between five Earths and 10 Jupiter masses," Cassan says. "If there are other planets farther or closer in, the average number of planets per star would increase accordingly."

The conclusion that smaller planets occur more often than bigger ones reinforces what Kepler has shown for planets that orbit close to their stars. The spacecraft is designed to locate worlds similar to our own—small, rocky planets at temperate, Earth-like distances from their host stars. That hunt is still underway, but early results from the mission have revealed that smallish planets—those just a bit bigger than Earth—are common in the hotter, close-in orbits to which Kepler is already sensitive.

"Kepler has already been finding that small planets are actually quite ubiquitous around stars," Gaudi says. "That bodes well for our goal of eventually finding an Earth-size planet in the habitable zone. All signs are pointing to the low-mass planets being common, so I think there's a good chance that we'll find a system like that in the coming years."