Elon Musk has spent two decades arguing that humanity must become a multiplanetary species, and that a self-sustaining city on Mars is the insurance policy for civilisation. The idea is stirring. It is also being sold as a matter of engineering and willpower, when the problems involved are mostly biological, chemical and planetary, and no amount of rocket thrust solves them. Here is what a serious colonist would face.
The core of the pitch is the "backup" argument: if an asteroid, pandemic or nuclear war ravages Earth, a Martian population survives. But compare the worst-case Earth with the best-case Mars. After the dinosaur-killing impact, the Earth still had breathable air, liquid water, a magnetic field and soil that grows crops. Mars has none of these. A post-apocalyptic Earth is a more hospitable place than a thriving Mars.
Early colonies would also depend entirely on Earth for spare parts, electronics, medicines and specialised machinery. A modern economy rests on millions of interdependent components, and estimates of the population needed for real self-sufficiency run from tens of thousands to millions. A lifeboat that can't be built or maintained without the ship it is meant to rescue isn't much of a lifeboat, at least not for a very long time.
Radiation is arguably the hardest problem because it is the least negotiable. Earth's magnetic field and thick atmosphere shield us from galactic cosmic rays, which are high-energy atomic nuclei, and from solar particle events. The average person receives about 2.4 millisieverts a year. Neither Mars nor the Moon offers anything close to that protection.
Transit to Mars: Instruments aboard Curiosity measured roughly 1.8 mSv per day in deep space, so a six-to-nine-month trip alone consumes a large fraction of NASA's career exposure limit for astronauts (600 mSv).
The Martian surface: Mars's thin atmosphere and lack of a global magnetic field leave the surface at roughly 0.6 mSv per day, on the order of 100 times Earth's average annual dose over a year.
The Moon: With no atmosphere at all, surface dose measured by China's Chang'e-4 lander is about twice Mars's, roughly 1.4 mSv per day.
Cosmic rays are hard to stop. Thin metal shielding can make things worse by generating secondary particles, so the practical answer is mass: living under several metres of regolith or inside lava tubes. That means a colony that lives underground, with radiation-hardened everything, and a real question about long-term cancer risk, central nervous system effects and fertility that nobody has data on, since no human has spent years beyond low Earth orbit's protective envelope.
Moon versus MarsThe Moon's dangers are harsh but near: extreme temperature cycling, vacuum, micrometeorites, abrasive dust that wrecked seals and irritated Apollo astronauts' lungs, and slightly higher radiation. Its big advantage is proximity. If a habitat fails, help and supplies are days away, and real-time control from Earth is possible.
Mars offers more of what a settlement needs, including some atmosphere for extracting oxygen and fuel, more accessible water ice, sunlight that is usable if weak, and a more favourable gravity. But its dangers are far: no abort option, no rescue, a months-long resupply cycle and global dust storms that can choke solar power for weeks. Perchlorate salts in the soil are toxic to humans, particularly the thyroid, which complicates both farming and dust exposure.
Then there is gravity. Nobody knows whether humans can develop normally, reproduce and raise children at 0.38 g, let alone 0.16 g. Astronauts in microgravity lose bone and muscle and suffer vision changes. Mammal-reproduction experiments in partial gravity have never been done at meaningful scale. Colonisation as a permanent, generational project rests on an unanswered biological question.
Musk has proposed warming Mars by detonating nuclear weapons over its poles, a line that made for headlines. The physics doesn't cooperate. A 2018 study by Bruce Jakosky and Christopher Edwards estimated that all the CO₂ accessible in the polar caps, regolith and minerals would raise the atmosphere to at most a few percent of Earth's surface pressure, far too little for a meaningful greenhouse effect, and that is before accounting for the difficulty of releasing it all. Much of the planet's original atmosphere was stripped away by the solar wind over billions of years, a process still measured today by NASA's MAVEN orbiter, and one that a rejuvenated atmosphere would suffer too, absent a magnetic field.
Even a heroic success would only raise pressure and temperature. Mars would still lack oxygen. A human-breathable atmosphere needs about 20% oxygen at meaningful total pressure, which means producing quantities of the gas that dwarf anything we've ever generated. And the timescale, even under optimistic scenarios, runs to centuries or millennia. Terraforming is a project for civilisations, not startups.
The Moon is worse still. Its low gravity gives an escape velocity of only about 2.4 km/s, so any atmosphere we added would bleed into space over comparatively short geological time. Lunar settlement means sealed habitats forever, not a garden world.
Realistic proposals shift toward "paraterraforming": enclosed, pressurised, radiation-shielded habitats on or under the surface. That is buildable, but it's a very different vision from a planet with blue skies.
There are further questions that engineering doesn't answer:
Planetary protection. If Mars hosts, or once hosted, microbial life, human arrival risks contaminating and destroying one of the most important scientific discoveries possible. The Outer Space Treaty requires avoiding harmful contamination, and enforcement is thin.
Governance. SpaceX's own terms of service declare that Mars is a "free planet" not subject to Earth governments, which sits awkwardly with the Outer Space Treaty's bar on national sovereignty claims. Who legally governs a settlement, and who is responsible when people die?
Consent and risk. The first settlers would face high odds of death, dependence on a single company for air and return transport, and very limited exit. That raises hard questions about informed consent, labour and coercion.
Opportunity cost. Billions spent on a Martian foothold are billions not spent on the fragile, valuable planet we already live on. This is not an either/or in principle, but resources and attention are finite.
None of this makes space settlement pointless. The best case for it isn't insurance against extinction. It is exploration, science, the technology spun off along the way, and a slow, patient expansion of what humans can do. The Moon is the logical first step, a nearby proving ground where we can test closed-loop life support, radiation shielding and in-situ resource use with a rescue option. Musk himself has more recently spoken of the Moon as a stepping stone, which suggests even the boldest advocates are recalibrating.
The honest framing is that Mars isn't a second Earth. It's an extraordinarily hostile frontier, more like Antarctica in a vacuum, and it may stay that way for a very long time. Treating it as humanity's destiny risks obscuring the near-term truth: the only planet we know how to live on is this one, and every challenge above is a reminder of how remarkable, and how worth protecting, it is.
https://www.youtube.com/watch?v=ax4wtOTaX7o