When will the first humans successfully land on Mars?

TL;DR

My median estimate is June 2044 UTC for the first successful human landing on Mars. The leading path is a Starship-derived system with NASA as customer or partner after several lunar and uncrewed Mars proving cycles. The distribution puts 0.2% before 2030, 5% by 2035, 23% by 2040, 59% by 2045, 76% by 2050, and 85.5% by 2056.

Context

As of July 2, 2026, I do not find a public, funded, dated, end-to-end crewed Mars landing program. NASA’s public path is still Moon first: Artemis II splashed down on April 10, 2026 after a nearly 10-day lunar flyby, Artemis III is a 2027 low-Earth-orbit rendezvous and docking test with commercial lander test articles, and Artemis IV is the first planned Artemis lunar surface landing in early 2028 (NASA Artemis II, NASA Artemis III, NASA Artemis IV). NASA’s Moon to Mars Architecture explicitly describes a blueprint for capabilities, not a mission, manifest, or fixed set of requirements (NASA Moon to Mars Architecture).

SpaceX remains the fastest visible path because Starship is designed to carry crew and cargo to Earth orbit, the Moon, Mars, and beyond, but Starship is still in flight test rather than operational Mars service (SpaceX Mars). As of late June 2026, Starship had not reached orbit and in-space refueling had not been demonstrated; Flight 12 on May 22, 2026 also led the FAA to require a mishap investigation of the Super Heavy booster (Space.com Starship status, FAA Flight 12 statement). China is a serious later competitor, with a crewed Moon landing target by 2030 and a Tianwen-3 Mars sample-return mission scheduled for launch around 2028 and return around 2031, but not a public crewed Mars landing program (China State Council, May 2026, China State Council, July 2025).

Evidence

The historical backbone is slow. First-of-kind human spaceflight systems usually take about a decade from commitment to a first crewed milestone, and Mars is a harder target than every row in this table except the political intensity of Apollo.

Mars adds a hard calendar constraint. Earth laps Mars about once every 26 months, so a mission that misses readiness by months often slips by roughly two years; recent robotic Mars cruises also show that touchdown comes months after launch rather than on the launch date (JPL Mars 2020 launch press kit). That is why my forecast is lumpy rather than smooth.

The bullish evidence is SpaceX’s industrial capacity. SpaceX’s June 2026 EU prospectus, with operating data through March 31, 2026, reports about 620 orbital launches, an over 99% mission success rate, 165 Falcon 9 launches in 2025, and about 7,400 metric tons launched to orbit over the company’s history (SpaceX EU prospectus). No state space program has a comparable visible launch cadence. If Starship matures, repeated cargo attempts to Mars become much more affordable than under any traditional NASA-only architecture.

The bearish evidence is that the mission-critical pieces are still open. NASA’s March 2026 HLS audit says SpaceX’s vehicle-to-vehicle cryogenic propellant transfer test had slipped 12 months to March 2026, that cryogenic propellant transfer is one of the most significant technical challenges for HLS, and that the processes being used are new and have never been done vehicle-to-vehicle (NASA OIG HLS audit). That is for a lunar lander. A first Mars landing also needs a crew-rated transit system, autonomous long-duration life support, surface power, Mars suits or at least safe egress, communications, and either ascent-return capability or a politically accepted no-return plan.

Mars entry, descent, and landing is the cleanest technical blocker. NASA’s 2024 Mars EDL white paper counts only 12 successful robotic landings out of 19 attempts, says all flown Mars EDL systems landed only 0.3 to 1 metric ton, and says human-class Mars EDL requires over 20 metric tons, more than 20 times the demonstrated class (NASA Mars EDL white paper). The same paper says there are no Earth-analog test conditions that fully mimic Mars EDL, so a Mars landing system cannot be validated with a true test-as-you-fly campaign on Earth (NASA Mars EDL white paper). NASA’s January 2026 civil-space shortfalls list keeps the same gaps alive: large-payload Mars landing, Mars ISRU, Mars ascent, low-loss cryogenic transfer at human-mission scale, long-duration habitation, radiation countermeasures, and Earth-independent medical care (NASA 2026 Civil Space Shortfalls).

NASA is now spending real money on Mars-relevant capabilities, but not enough to treat a landing as scheduled. The FY2027 request includes a Mars Technology line of $438.8 million in FY2027 rising to $680.2 million in FY2030, plus commercial Moon and Mars infrastructure and transportation lines, but it does not specify a crewed Mars landing date or crewed Mars landing vehicle procurement (NASA FY2027 budget request). I read this as serious enabling investment, not a mission of record.

China mostly affects the later half of the distribution. Its crewed lunar program is concrete: the May 2026 CMSA release says China is targeting a crewed lunar landing by 2030, has tested Long March-10 and Mengzhou systems, and still needs the Long March-10 technical verification flight plus maiden flights of Mengzhou and the Lanyue lunar lander (China State Council, May 2026). Its Tianwen-3 plan is also concrete and Mars-relevant, because it requires Mars sampling, takeoff, rendezvous in Mars orbit, and planetary-protection controls, but it is robotic and targets launch around 2028 and sample return around 2031 (China State Council, July 2025). A 2025 China Aerospace Studies Institute report argues for a Chinese crewed Mars orbit mission around 2050 and explicitly notes that this is not the same as a landing (CASI report).

I modeled the resolution date as a mixture over Mars arrival opportunities, with a small pre-2030 tail and a structural post-2056 tail. The finite windows are centered on approximate arrival opportunities from 2031 to 2055 and spread with an 85-day normal kernel: ( P(T \le t)=p_{<2030}+\sum_i w_i\Phi((t-\mu_i)/85\text{ days}) ), where $\mu_i$ is a landing-window center and $w_i$ is my assigned probability mass for that window. I assign 0.2% before 2030, 85.3% across windows before 2056, and 14.5% after 2056. The window weights peak in 2040 to 2044, where a successful SpaceX path, a NASA-commercial path after Artemis, and a Chinese pressure path first overlap. The resulting cumulative probabilities are 4.7% by 2035, 22.9% by 2040, 59.1% by 2045, 76.2% by 2050, and 85.5% by 2056.

What's non-obvious

The obvious forecast is to split the difference between SpaceX’s old public dates and NASA’s slower rhetoric. That frame misses the bottleneck. Launch cadence is no longer the whole problem; the hard part is turning a heavy launcher into a human Mars mission system. Orbital refueling, Mars EDL, surface power, crew health, and mission operations under communication delay are independent gates. Starship can compress several of them, but it has not yet compressed all of them.

The Moon program is both accelerator and bottleneck. Artemis and a Moon-first SpaceX path should produce refueling, lander operations, surface power, suits, construction, and life-support lessons that help Mars. They also consume the next several years of engineering attention. A lunar stack that still needs a 2027 low-Earth-orbit test before an early-2028 landing does not support a normal-case human Mars landing in the early 2030s.

Limitations

The largest uncertainty is private SpaceX execution and intent. Public sources show flight tests, NASA contracts, and regulatory events; they do not show internal Mars Starship design maturity, life-support work, crew-risk tolerance, or how much capital and talent SpaceX will devote to Mars while serving lunar, Starlink, and other commercial priorities. A successful uncrewed Starship Mars landing in the 2028 or 2031 window would move substantial probability from the 2040s into the mid-to-late 2030s.

China is also opaque. Its Moon and Tianwen-3 plans are public enough to model as real precursors, but a crewed Mars landing would require a program the public record has not yet shown. A formal Chinese crewed Mars landing approval after a successful lunar landing would raise the 2045 to 2056 mass.

The model also assumes the first attempt is made by a state or state-backed company with a strong incentive to avoid killing the first Mars crew. The resolution criteria only require one human to survive touchdown for a non-zero period, so a high-risk one-way or weak-return mission is not impossible. I include that mainly in the small 2031 to 2038 mass rather than treating it as the central path.