Posted by There is a difference between visiting a place and living there, and most plans to colonize Mars blur it. Planting a flag, spending a month doing geology and flying home is an expedition. A colony is something else: people who stay, have children, fix their own machines and eventually stop needing a cargo ship from Earth every 26 months. Nobody has a working plan for the second thing yet. What exists is a stack of proposals, a handful of real experiments and a long list of problems that are only partly solved.
Big promises, small hardware
The loudest voice belongs to SpaceX. Elon Musk has talked for years about a self-sustaining city on Mars, at one point putting the target population at a million people, carried there by fleets of Starship vehicles that refuel in Earth orbit and again on the Martian surface. Starship has flown a long series of test flights from Texas, with spectacular failures alongside genuine progress, but it has not yet carried a person or landed anything on another planet. Dates for an uncrewed Mars attempt have been announced and quietly moved more than once.
Government plans are more modest and slower. NASA frames Mars as the destination after Artemis lunar missions, with crewed flights vaguely placed somewhere in the 2030s or beyond, and its documents talk about expeditions and a possible base, not a colony. China has outlined its own crewed Mars ambitions on a similar horizon. None of these organizations has a funded program for a permanent settlement. That word, permanent, is doing an enormous amount of work in public discussion.
Older blueprints still shape the debate, and almost all modern plans to colonize Mars lean on one of them. Robert Zubrin’s Mars Direct plan from 1990 argued that you could make the return fuel on Mars rather than haul it from Earth, which turned the whole problem from “impossibly heavy” into “merely very hard”. Most serious architectures since then borrow that idea.
Living off the land, one gram at a time
Making things on Mars, rather than shipping them, is the core of every settlement concept. Engineers call it in-situ resource utilization, and the honest status report is: one small proof of concept, done well.
That proof was MOXIE, a toaster-sized instrument on the Perseverance rover. Between 2021 and 2023 it pulled carbon dioxide from the Martian air and split it into oxygen sixteen times, producing a total of about 122 grams. Its best runs made roughly 12 grams an hour, about what a small dog breathes. It worked in different seasons, day and night, which is the real headline. Scaling it up by a factor of hundreds to supply both breathing air and the oxygen for rocket propellant is an engineering job nobody has done yet.
Water is the bigger prize. Radar from orbiters such as Mars Reconnaissance Orbiter and Mars Express has mapped buried ice across the planet, and fresh impact craters have exposed bright water ice in the mid-latitudes, sometimes just below the surface. A settlement would almost certainly sit somewhere like that, perhaps in a region such as Arcadia Planitia, close enough to the equator for decent sunlight and close enough to the ice to dig. With water and carbon dioxide you can, in principle, make methane and oxygen through the Sabatier process, which is exactly the propellant Starship uses. In principle. The plant to do it at scale has never been built off Earth.
The soil, the sunlight and the dust
The ground itself is unfriendly. NASA’s Phoenix lander found perchlorates in Martian soil in 2008, and later missions confirmed they are widespread. These salts are toxic to people at the concentrations found there, which means every handful of regolith brought into a greenhouse has to be washed first, and every speck of dust tracked into a habitat on a spacesuit is a slow health problem. Lunar dust gave Apollo astronauts irritated eyes and congested noses after just a few days. Mars dust is finer, chemically nastier and everywhere.
Power is another headache. Sunlight at Mars is a little under half as strong as at Earth, and global dust storms can dim it for weeks. The Opportunity rover died that way in 2018 after nearly fifteen years of work. A colony running only on solar panels would be one bad storm season away from rationing heat and air, so most designs pair solar arrays with compact nuclear fission reactors.
Food is a quieter, harder problem
Grow your own. Easy to say. Plants need pressure, warmth, light, water and nutrients, and every one of those has to be manufactured or recycled inside a sealed building. Experiments on the International Space Station have grown lettuce, radishes and chili peppers in small chambers, and researchers in the Netherlands have coaxed crops from simulated Martian soil once the toxic chemistry was dealt with. Feeding a hundred people, year after year, with no resupply, is a different scale entirely.
The cautionary tale here is Biosphere 2 in Arizona. In 1991 eight people sealed themselves inside a closed ecosystem for two years. Oxygen levels dropped so far that fresh air eventually had to be pumped in, crops underperformed and the crew lost a lot of weight. It was in Arizona, with sunlight and a phone line. It is still the best real-world reminder of how hard closed ecosystems are to balance.
How many people make a colony?
This is where the engineering questions become human ones. A self-sustaining settlement needs not just food and air but people who can repair a pump, deliver a baby, run a chip fab or at least replace one. Estimates of a minimum viable population range wildly, from a few hundred to many thousands, depending on how much you assume can be automated or shipped in. Nobody really knows.
Then there’s gravity. Mars has about 38 percent of Earth’s surface gravity, and we have essentially no data on what years of that does to bones, muscles, eyes or pregnancy. Studies from the space station describe microgravity, which is a different situation. Whether children could develop normally on Mars, or whether people raised there could ever comfortably visit Earth, are open questions. Not unanswered because nobody asked. Unanswered because there is no way to test them except by going.
Law and politics trail behind. The 1967 Outer Space Treaty forbids any nation from claiming sovereignty over a celestial body, but says little about who governs a settlement run by a company, or who owns the water it mines.
What a first settlement would actually look like
Strip away the renderings and the most plausible first step is a research station, closer in spirit to the Amundsen–Scott base at the South Pole than to a city. A few dozen people at most, living in landed modules partly buried under regolith for radiation shielding, with a pilot propellant plant, a reactor, a greenhouse and a lot of spare parts. Crews would rotate. Much of the equipment would come from Earth on every launch window, which opens roughly every 26 months when the two planets line up.
From there, the path to a true colony goes through boring milestones: making more of your own air than you import, then water, then propellant, then spare parts, then food. Each one cuts the umbilical a little. The hardest might be the last. Manufacturing a single modern microchip relies on supply chains spread across dozens of countries.
So how close are the plans to colonize Mars? Closer than in Zubrin’s day, thanks to reusable rockets, better ice maps and a working oxygen machine on a rover. Still far from a place where anyone could reasonably expect to grow old.
If questions like these are what keep you up at night, SETIworld has plenty more on Mars, habitability and humanity’s chances beyond Earth. Read on, join the discussion, and bring your own estimate for that minimum viable population.
