Twenty-six years of continuous presence
The International Space Station has been continuously crewed since 2 November 2000, the longest period of uninterrupted human presence in space. Assembly began in 1998 and took more than a decade of Space Shuttle, Proton and Soyuz flights. The completed station has about 916 cubic metres of pressurised volume, of which about 388 cubic metres is habitable, a mass of about 420 tonnes and a crew of seven on six-month rotations. Five agencies share it under the 1998 Intergovernmental Agreement: NASA, Roscosmos, ESA, JAXA and the Canadian Space Agency.
The station's scientific record is substantial. thousands of investigations across biology, human physiology, physical sciences, Earth observation and technology demonstration, and a US National Laboratory that has hosted commercial and academic research since 2011. Its cost is also substantial. Public estimates put the station's total cost above $150 billion, about two-thirds of it funded by NASA. NASA's Office of Inspector General found in 2024 that the ISS and related operations and research cost about $4.1 billion a year, around 16% of NASA's budget in fiscal 2023 and 2024. A 2026 Government Accountability Office report put the cost of maintaining and operating the station at about $3 billion a year, of which crew and cargo transport is about 60%.
The decision to retire
NASA plans to end ISS operations in 2030 and to deorbit the station in a controlled re-entry over the remote South Pacific. The decision rests on three arguments. The structure is ageing: the oldest modules have been in orbit since 1998 and 2000, and every year adds fatigue cycles and maintenance cost; the Inspector General found in 2021 that systems maintenance and upgrade costs had risen by 35% over five years. The partnership is under strain: Russia's commitment beyond the late 2020s is uncertain, and the Russian segment provides propulsion and attitude control the US segment depends on. And NASA wants to redirect the station's budget towards the Artemis lunar programme and Mars.
In June 2024 NASA selected SpaceX to build the US Deorbit Vehicle, with a potential contract value of $843 million excluding launch. Unlike Dragon, which SpaceX owns and operates, NASA will own and operate the deorbit vehicle once development ends. It is due to be ready in 2029. The deorbit contract turns the end of the ISS from an aspiration into a scheduled engineering event: the vehicle needs a free docking port, a burn plan and a launch, and those constraints now shape other programmes, including Axiom's assembly sequence.
Why an extension is unlikely
The ISS has been extended before, from 2015 to 2020, to 2024 and to 2030. A further extension would require new appropriations, a new agreement with the partners, confidence in the structure and a revised deorbit plan. The structural question is the hardest. A slow air leak in the PrK transfer tunnel of the Russian Zvezda module has been monitored for several years; repairs reduced but did not stop it, and in 2021 the Inspector General reported an air loss rate still double the baseline. NASA and Roscosmos have both treated it as a top programme risk. Several commentators have noted that a new leak could force questions about the station's life before 2030, not after. On balance an earlier retirement is more plausible than a later one, which shortens rather than lengthens the window for commercial successors.
The transition logic
NASA's strategy since its 2019 commercial LEO directive has been to stop owning and operating a station and to buy services instead, as it does for crew and cargo transport. Three arguments support it. NASA wants to redirect the $3–4 billion a year the ISS absorbs towards exploration. A commercial market, with more than one customer, should lower the unit cost of research time. And a US-led presence in orbit is a strategic objective in its own right while Tiangong operates. The model is commercial crew and cargo, where NASA paid part of development through milestones and then bought services under fixed-price contracts, leaving ownership with the provider. The difference is that crew and cargo had one obvious customer and a fixed demand; a station's economics depend on filling it.
Phase 1: designs and an ISS-attached module
The first phase ran from 2020. Axiom Space won a firm-fixed-price contract in January 2020, worth $140 million, to attach commercial modules to the ISS. In December 2021 NASA signed funded Space Act Agreements for free-flying station designs with Blue Origin (Orbital Reef, $130 million), Nanoracks with Voyager and Lockheed Martin (Starlab, $160 million) and Northrop Grumman ($125.6 million). Northrop Grumman later withdrew its own concept and joined the Starlab team, whose agreement was increased to $217.5 million. These awards paid for design milestones, not stations. NASA's total development funding across all providers remains under $600 million, and in fiscal 2021 Congress appropriated $17 million for the programme against a request of $150 million.
Phase 2: fourteen months of change
| Date | Step | What changed |
|---|---|---|
| Aug 2025 | NASA directive | Phase 2 moved from fixed-price contracts to funded Space Act Agreements, branded C3DO, with $1.0–1.5 billion over fiscal 2026–2031 and at least two awards; the requirement fell from continuous crewed presence to four crew for 30-day increments |
| 28 Jan 2026 | Pause | Phase 2 placed on hold pending a strategy review by the new Administrator, Jared Isaacman, sworn in on 18 December 2025 |
| 24 Mar 2026 | "Ignition" | NASA proposed buying a government-owned core module with two docking ports, attached to the ISS, to which commercial modules would dock before the stack flew free |
| 1 Jun 2026 | Reversal | After objections from the industry, represented by the Commercial Space Federation's seven station companies, NASA withdrew the core-module plan and returned to commercially owned stations |
| Jul 2026 | Draft RFP | Firm-fixed-price, multi-award, indefinite-delivery contracts for development, certification and services; industry comments due late July |
| 9 Oct 2026 | Final RFP | Solicitation 80JSC027R0004; proposals due 8 December 2026; awards expected in spring 2027 |
The sequence matters for anyone underwriting these programmes. The March proposal signalled that NASA was prepared to own core infrastructure if it doubted industry could deliver on time, and it favoured the architecture closest to it (Axiom's ISS-attached modules) over free-flyers (Starlab, Vast's Haven-2, Orbital Reef). The June reversal restored the commercial model, and Administrator Isaacman framed the July draft around "a viable commercial marketplace" in which "NASA is one customer among many". But the episode is now in every provider's risk register: the requirements, the contract vehicle and the architecture have all moved in the space of a year.
The final procurement
The final request for proposals, a 360-page solicitation, asks companies to design, build, test, certify and operate commercial destinations providing end-to-end services for NASA's human spaceflight needs. Its structure has three steps. NASA intends to award early development contracts to at least two providers. It will then run a competition among them for a task order covering final design, test, evaluation and certification, which may go to one or more. Services task orders follow certification. The contracts are firm-fixed-price and indefinite-delivery: NASA pays for milestones and services delivered, and providers bear the cost of overruns.
Reported figures from the solicitation include a guaranteed minimum value of $100 million and a ceiling of about $34 billion over the contract's life, together with NASA planning prices of about $325 million for a four-seat crew flight and $300 million for a cargo flight in 2030. The ceiling is an upper bound for the contract vehicle, not a budget: actual funding depends on annual appropriations. The planning prices matter for station economics because they set the cost of getting NASA's crews and supplies to whichever station it buys from.
The money behind the transition
For fiscal 2027 the administration requested $18.8 billion for NASA, a 23% reduction on current enacted levels. Within it, the request for commercial LEO development is $299.7 million, up $26.7 million or 9.8% on fiscal 2026, while ISS funding falls to $912.2 million, a cut of about a quarter. The House Appropriations Committee rejected the deep cuts: its bill provides no less than $400 million for commercial LEO development, $1.49 billion for the ISS with direction to maintain the fullest possible utilisation, and $1.344 billion for commercial crew and cargo. It also asked NASA to report on the acquisition schedule and to ensure that at least one commercial destination is available by the end of ISS operations. The final appropriation was not enacted at the time of writing.
Either figure is small against the cost of a station. The transition is being financed mostly by private capital on the expectation of NASA services revenue after 2030, which makes the stability of NASA's procurement and the scale of its services budget the two variables that matter most to the providers' financing.
Transport to and from the stations
A station is only as useful as the vehicles that reach it. SpaceX's Crew Dragon, certified in November 2020, is the only operational US crew vehicle. Boeing's Starliner has not been certified: its June 2024 crew flight test was classified as a Type A mishap, NASA's highest category, the astronauts returned on Dragon, and NASA now plans an uncrewed Starliner-1 flight around the turn of 2026–27 and a return to crewed flights by 2028. Cargo is more diverse: SpaceX's Cargo Dragon, Northrop Grumman's Cygnus (Cygnus XL flew NG-24 on a Falcon 9 in April 2026), JAXA's HTV-X (first flight October 2025) and, prospectively, Sierra Space's Dream Chaser, whose first flight is now a free-flying demonstration no earlier than late 2026. ESA's LEO cargo return programme has moved its first demonstration to a 2029 ISS flight and will accept missions to commercial destinations. Dependence on one crew vehicle is a risk every station shares.
The gap risk
If no commercial station is ready to host NASA crews when the ISS retires, the United States faces a gap in continuous human presence in orbit for the first time since 2000. The consequences are practical and political. Research programmes pause and their teams disperse. Astronaut long-duration proficiency and the medical knowledge built on it erode. The supplier base for station hardware, already shrinking as the ISS approaches its end, loses customers. International partners look elsewhere. And Tiangong becomes the only permanently crewed station, with its own offer of access to the countries that would otherwise fly on Western stations. The interactive model in chapter 4 shows how many free-flying stations are ready under different schedule slips.