The $11 Billion Tube

Somewhere on the floor of Jezero Crater, a titanium tube the size of a dry-erase marker holds a rock core that may contain evidence of ancient Martian life. The tube is called Sapphire Canyon. It was drilled from a rock called Cheyava Falls by the Perseverance rover on July 21, 2024, hermetically sealed, and stored in the rover’s belly carousel alongside 32 other sample tubes—the most scientifically valuable collection ever assembled off Earth.¹ As of February 2026, no funded program exists to bring it home.

This is the story of Mars Sample Return: how a $2.5 billion concept became an $11 billion program, how the program died in the gap between scientific consensus and congressional appropriation, and what paths remain.

The Idea That Wouldn’t Die

Mars sample return is not a new idea. It is arguably the oldest unfulfilled ambition in planetary science. The concept appeared in NASA planning documents as early as 1970, when the post-Apollo glow still suggested that bringing rocks home from other planets was a thing the United States simply did.²

It surfaced again in 1996, when the Allan Hills 84001 meteorite controversy electrified the field. A team led by David McKay at Johnson Space Center published a Science paper claiming to have found nanoscale fossil structures in a Martian meteorite that had been sitting in an Antarctic ice field for 13,000 years. President Clinton held a press conference. The National Academy of Sciences convened a review. The putative nanofossils turned out to be ambiguous—too small to be terrestrial cells, too morphologically simple to distinguish from mineral artifacts. The debate was never definitively resolved.³

But ALH 84001 demonstrated a principle that would drive the next three decades of Mars exploration: you cannot answer the life question with instruments on Mars. You need samples on Earth, in laboratories with instruments that are too heavy, too complex, and too power-hungry to fly. The meteorite had been analyzed with transmission electron microscopy, ion microprobe mass spectrometry, laser Raman spectroscopy, and energy-dispersive X-ray spectroscopy—a combined analytical suite weighing tens of thousands of kilograms and requiring stable environments, controlled atmospheres, and human operators. No rover will carry that. The answer, if it exists, requires a round trip.

Three Decadal Surveys, One Recommendation

Planetary science in the United States is governed by decadal surveys—massive community assessments conducted by the National Academies every ten years, ranking scientific priorities and recommending missions. They are the closest thing the field has to a democratic mandate. NASA is not legally bound by them, but defying a decadal survey costs political capital that administrators rarely want to spend.

Mars sample return has been the top priority of three consecutive decadal surveys:

2003 — “New Frontiers in the Solar System” Recommended MSR as the highest-priority large mission for the decade. Estimated cost: not formally baselined. 2011 — “Vision and Voyages” Reaffirmed MSR as the top flagship priority. Baselined the Mars 2020 rover (now Perseverance) as the caching step. Estimated MSR campaign cost: ~$2.5 billion beyond Mars 2020.⁴ 2023 — “Origins, Worlds, and Life” Named MSR the highest scientific priority in all of planetary science. Stated: “MSR is critical to addressing whether Mars has ever harbored life.” Recommended a cost cap of $5.3 billion for the retrieval and return phases.⁵

Three surveys across two decades. The scientific community spoke with unusual unanimity: this is the most important thing we can do in planetary science. Every rover, every orbiter, every lander sent to Mars since 2003 was, in some sense, a precursor to the sample return that the community believed was coming.

Perseverance itself was designed as the first step. Its primary mission is not to study Mars. Its primary mission is to collect, seal, and cache samples for a future retrieval mission. Every drill site, every core, every sealed tube is predicated on the assumption that something would come to pick them up. The rover is, architecturally, a packing facility for a shipping service that was never contracted.

The Architecture

Mars Sample Return, as designed through 2023, was a three-mission campaign involving two space agencies and four spacecraft:

Mission 1: Mars 2020 (NASA) — Launched July 2020, landed February 2021. Perseverance collects and caches samples. This mission is complete and operating. 33 tubes filled. 10 duplicate tubes deposited at the Three Forks sample depot as backup.¹

Mission 2: Sample Retrieval Lander (NASA/ESA) — A lander carrying two payloads: a Sample Fetch Rover (ESA) to drive to Perseverance or the Three Forks depot, collect tubes, and return them to the lander; and a Mars Ascent Vehicle (NASA/JPL)—a two-stage solid-fuel rocket that would launch the sample container into Mars orbit. The MAV would be the first rocket ever launched from the surface of another planet.⁶

Mission 3: Earth Return Orbiter (ESA) — An ESA spacecraft that would rendezvous with the sample container in Mars orbit, capture it, seal it in a secondary containment system, and fly it back to Earth. The ERO uses solar-electric propulsion for the return trip. Transit time: approximately 13 months. The samples would re-enter Earth’s atmosphere in a capsule modeled on the Stardust sample return system.⁷

On Earth: Sample Receiving Facility — A BSL-4-equivalent containment laboratory, not yet designed, not yet sited, not yet funded. The facility must contain Martian material under the assumption that it could be biologically active, while simultaneously allowing scientists to perform the analytical work—IRMS, NanoSIMS, TEM, synchrotron X-ray diffraction—that the samples were collected to enable. No such facility exists. The closest analogue is the Lunar Sample Laboratory at Johnson Space Center, which handles Apollo samples under nitrogen atmosphere but without biological containment. The Mars facility would need both.⁸

The architectural elegance is real. Four spacecraft, two agencies, three planetary surfaces (Earth, Mars, Mars orbit), one round trip. It is the most complex robotic mission ever conceived. It is also, as the review boards discovered, the most expensive.

The Escalation

Cost growth is endemic to large NASA missions. The James Webb Space Telescope was estimated at $1 billion in 1997 and delivered for $10 billion in 2021. Mars Science Laboratory (Curiosity) was estimated at $1.6 billion and delivered for $2.5 billion. Cost escalation in flagship missions is not the exception; it is the base case.⁹

MSR followed the pattern, but faster:

2011 Decadal survey baselines MSR retrieval at ~$2.5B (beyond Mars 2020’s $2.7B cost). 2020 NASA/ESA joint concept review estimates $3.8–$4.4B for retrieval and return phases. 2022 Internal NASA estimates rise to ~$5.3B. Schedule slips from 2031 return to 2033. September 2023 Independent Review Board reports: cost has grown to $8–11B. Return date has slipped to 2038–2040. The board finds “unrealistic budget profiles” and “lack of an adequate technical baseline.”¹⁰ March 2024 NASA Administrator Bill Nelson announces the agency will seek “innovative, out-of-the-box” new designs. The existing architecture is effectively shelved. JPL’s MSR team begins downsizing.¹¹ October 2025 JPL announces 530+ layoffs, the largest workforce reduction in the lab’s history. Many affected staff worked on MSR-related programs.¹² January 2026 Congressional minibus spending bill zeroes MSR line items. The Trump administration’s FY2026 budget request had already called MSR “financially unstable.” The program is effectively cancelled.¹³

From $2.5 billion to $11 billion in twelve years. From a 2031 return to a 2040 return. From the top priority in planetary science to a zeroed budget line. The trajectory is not unusual for a NASA flagship. What is unusual is that the scientific case strengthened as the program collapsed. Perseverance kept drilling. The tubes kept filling. And in September 2025, four months before Congress killed the funding, the Cheyava Falls team published the strongest evidence yet that one of those tubes might hold the answer to the oldest question in astrobiology.¹⁴

The Pivot That Killed the Program

The Independent Review Board’s September 2023 report was damaging but not fatal. The board recommended restructuring, not cancellation. Their proposed path: descope the Sample Fetch Rover, rely on Perseverance to deliver tubes directly to the lander (the rover was still healthy), retain ESA’s Earth Return Orbiter, and simplify the Mars Ascent Vehicle. Estimated cost of the restructured mission: approximately $5.3 billion. Return date: 2035.¹⁰

NASA rejected the restructure. In March 2024, rather than committing to the $5.3 billion path, the agency issued a call for “new and innovative mission designs.” Eleven proposals were submitted from NASA centers, commercial partners, and international agencies. A study period of 12–18 months was announced.¹¹

This decision is the hinge point. The IRB had given NASA a buildable, costed, descoped architecture with a credible schedule. NASA chose to study alternatives instead. Study cycles in aerospace run 18–24 months from solicitation to downselect. Congressional appropriation cycles run 12 months. The study was structurally incapable of producing a buildable design before the next budget window opened—and in that window, MSR had no design to defend, no cost estimate to cite, no schedule to commit to. It was a program in search of an architecture, asking Congress to fund the search.

Congress declined.

Whether NASA’s decision was a genuine attempt to find a cheaper path or a bureaucratic mechanism to kill a program that had become politically untenable is a question that the participants answer differently depending on which building they work in. What is not in dispute is the outcome: the study period outlived the funding window, and when the January 2026 appropriation came, there was nothing concrete to appropriate for.

The Irreversible Loss

Budget lines can be restored. Workforce cannot—not at the same cost, not with the same institutional knowledge, not on the same timeline.

JPL’s October 2025 layoffs cut 530 positions, approximately 8% of the lab’s workforce. This followed an earlier round of 100 contractor terminations in February 2025 and the non-renewal of several hundred term positions throughout the year. The cumulative reduction since the MSR pivot exceeded 800 people.¹²

These are not generic engineering positions. Mars sample handling, autonomous rendezvous in planetary orbit, solid-fuel ascent vehicle design for non-Earth launch, planetary protection containment protocols—these are specializations with talent pools measured in dozens, not thousands. The engineers who designed Perseverance’s sample caching system, who prototyped the Mars Ascent Vehicle’s ignition sequence for Martian atmospheric pressure, who wrote the guidance algorithms for orbital rendezvous around a planet with a 24-minute communication delay—when they leave JPL, they take knowledge that is not written down, because it was never finished.

ESA’s situation is different but not better. The Earth Return Orbiter is ESA’s contribution to MSR and one of the most technically ambitious spacecraft the agency has designed. Development continued through 2025 under existing ESA contracts, but without a NASA retrieval mission to provide the sample container in Mars orbit, ERO has no payload. ESA has signaled willingness to continue development if a retrieval path materializes, but the spacecraft cannot justify its own funding indefinitely without a partner mission to complete the campaign.⁷

The Other Program

While MSR collapsed, China’s Mars sample return program advanced.

Tianwen-3 is a two-launch mission planned for 2028. The first launch delivers a lander and ascent vehicle to the Martian surface. The second launch delivers an orbiter and return capsule. The lander collects samples, the ascent vehicle lofts them to orbit, the orbiter captures and returns them to Earth. Estimated return: 2031.¹⁵

The architecture is simpler than NASA’s MSR design. No fetch rover. No pre-cached samples. Tianwen-3 collects its own samples at its own landing site—likely in Utopia Planitia, where the Tianwen-1 rover Zhurong operated in 2021–2022. It would not visit Jezero Crater. It would not retrieve Perseverance’s tubes. The Cheyava Falls sample would remain on Mars.

But Tianwen-3 would accomplish something with enormous downstream consequences: it would make China the first nation to return samples from Mars, and it would require China to build the receiving infrastructure—the containment laboratories, the analytical suites, the planetary protection protocols—that the United States has not yet built because MSR was always five years away.

China already has operational sample return infrastructure. The Chang’e 5 mission returned 1,731 grams of lunar material in December 2020. Chang’e 6 returned samples from the lunar far side in June 2024—the first mission ever to do so. The sample handling, re-entry, and curation pipeline is operational. Scaling it from lunar (no biological containment required) to Martian (full containment required) is a significant engineering challenge, but the organizational capability exists.¹⁶

The geopolitical framing is obvious and probably overdrawn. Mars sample return is not an arms race. There is no strategic advantage to possessing Martian regolith. But the infrastructure for analyzing extraterrestrial samples—the labs, the instruments, the trained personnel, the containment protocols—is a capability that, once built, defines who can participate in the next phase of planetary science. If China builds it first, the terms of international collaboration shift.

What Remains

MSR is not architecturally dead. It is fiscally suspended. The distinction matters because the hardware is at different stages:

Perseverance is operating. The sample collection is nearly complete. The rover’s nuclear power source (a multi-mission radioisotope thermoelectric generator) has a design life well beyond 2030. The samples in the Three Forks depot are deposited and stable. This phase of the campaign is done.¹

ESA’s Earth Return Orbiter is in development. The solar-electric propulsion system, the orbital capture mechanism, and the Earth re-entry capsule are under contract with Airbus Defence and Space. ESA has not cancelled ERO, though its continuation depends on a viable retrieval partner.⁷

The Sample Retrieval Lander and Mars Ascent Vehicle are cancelled. The MAV was the highest-risk element—a solid-fuel rocket designed to launch from the Martian surface at temperatures as low as −20°C, a scenario with no flight heritage. Significant testing had been completed, including hot-fire tests of the first stage, but the vehicle was years from flight readiness.⁶

The Sample Receiving Facility was never started. No site selected, no design finalized, no containment protocols certified. This is perhaps the most underappreciated gap: even if a retrieval mission launched tomorrow, there is no laboratory on Earth approved to open Martian samples under biological containment. Building one takes a minimum of 7–10 years from authorization to operation.⁸

Several alternative paths have been discussed:

Commercial retrieval. During NASA’s 2024 study period, proposals from commercial partners (including concepts leveraging Starship’s payload capacity) were submitted. None were selected. The technical challenge is not payload capacity but precision: landing within driving distance of Perseverance or the Three Forks depot, operating a sample transfer mechanism, and launching an ascent vehicle from the surface. These are not problems that scale with rocket size.

Perseverance self-delivery. If the rover remains healthy, it could potentially drive to a future lander and deliver tubes directly, eliminating the fetch rover. This was the IRB’s recommended descope. It depends on Perseverance surviving until a retrieval lander arrives—a bet on hardware longevity that gets riskier with every year of delay.

International partnership beyond ESA. Japan’s MMX mission (Martian Moons eXploration), launching in 2026, will return samples from Phobos by 2031. It will not visit the Martian surface, but its sample return capsule and Earth re-entry technology could inform a future MSR design. JAXA has expressed interest in broader Mars sample return collaboration.¹⁷

A future administration. MSR has survived three presidential administrations and two cancellation attempts. The program has institutional momentum within the planetary science community, decadal survey backing, and now—for the first time—a compelling in-situ finding to justify the expense. A future Congress could restore funding. The question is whether the workforce, the ESA partnership, and the hardware development will still be available when that happens.

The Tube

Titanium does not corrode in the Martian atmosphere. The sample tubes are hermetically sealed with indium metal gaskets. The cores inside are protected from UV radiation, temperature cycling, and atmospheric contamination. Perseverance’s engineers designed these tubes to preserve sample integrity for decades—long enough for a retrieval mission that, at the time of design, everyone assumed would come.

The tubes will be there in 2040. They will be there in 2060. The Martian surface is geologically stable on human timescales. Aeolian processes—wind-driven sand transport—could partially bury the Three Forks depot over decades, but the tubes are detectable by orbital radar and their GPS-equivalent coordinates are known to sub-meter precision.

The samples are patient. The institutions are not. JPL’s workforce is dispersing. ESA’s ERO has a finite development window. The planetary science community that built three decadal surveys around this mission is aging. The engineers who know how to build a Mars Ascent Vehicle are finding other jobs.

Sapphire Canyon sits in Perseverance’s sample carousel, tube number 24 of 43 slots, sealed and waiting. The rock inside may contain the mineral-organic associations that, on Earth, are overwhelmingly produced by life. The instruments to confirm that hypothesis are in laboratories in Houston, Pasadena, Grenoble, and Tokyo. The distance between the sample and the instruments is 140 million miles at closest approach. The distance between the scientific consensus and the political will to close that gap is, at present, infinite.

Part 3 examines the ocean worlds pipeline—Europa Clipper, Enceladus Orbilander, and the instruments being designed to detect life in subsurface oceans that have never seen sunlight.

---

Footnotes

1. NASA JPL. “The 33 Sample Tubes Collected by Perseverance.” jpl.nasa.gov; NASA. “Perseverance’s Three Forks Sample Depot Map.” science.nasa.gov ↩ ↩² ↩³

2. National Academies. “Assessment of Mars Science and Mission Priorities.” 2003. nap.nationalacademies.org ↩

3. McKay, D. et al. “Search for Past Life on Mars: Possible Relic Biogenic Activity in Martian Meteorite ALH84001.” Science, 1996. science.org ↩

4. National Academies. “Vision and Voyages for Planetary Science in the Decade 2013–2022.” 2011. nap.nationalacademies.org ↩

5. National Academies. “Origins, Worlds, and Life: A Decadal Strategy for Planetary Science and Astrobiology 2023–2032.” nap.nationalacademies.org ↩

6. NASA JPL. “Mars Ascent Vehicle.” jpl.nasa.gov ↩ ↩²

7. ESA. “Earth Return Orbiter.” esa.int; Airbus. “Mars Sample Return.” airbus.com ↩ ↩² ↩³

8. National Academies. “Assessment of Planetary Protection Requirements for Mars Sample Return Missions.” nap.nationalacademies.org ↩ ↩²

9. Government Accountability Office. “NASA: Assessments of Major Projects.” GAO-24-106015. gao.gov ↩

10. NASA. “Mars Sample Return Independent Review Board Report.” September 2023. nasa.gov ↩ ↩²

11. NASA. “NASA Seeks New Ideas for Mars Sample Return.” April 2024. nasa.gov ↩ ↩²

12. SpaceNews. “JPL to lay off about 530 employees.” spacenews.com ↩ ↩²

13. Scientific American. “NASA’s Mars Sample Return Mission in Jeopardy as U.S. Considers Abandoning It.” scientificamerican.com ↩

14. Sharma, S. et al. “Redox-driven mineral and organic associations in Jezero Crater, Mars.” Nature, September 2025. nature.com ↩

15. Nature. “China’s ambitious plan to bring back Mars samples.” nature.com ↩

16. Nature. “Chang’e 5 lunar samples.” nature.com; Nature. “Chang’e 6 returns lunar far-side samples.” nature.com ↩

17. JAXA. “Martian Moons eXploration (MMX).” isas.jaxa.jp ↩