SpaceX shifts from Falcon 9 to Starship as launch crunch looms

SpaceX is ending Falcon 9 operations and shifting its launch business to Starship, industry observers said at Silicon Valley Space Week on Oct. 27, expecting less capacity and higher prices across the launch industry.

Elon Musk said in August that Falcon would wind down once Starship flies reliably several times a week, setting a reliability condition rather than a fixed date for the retirement. NASA Administrator Jared Isaacman said on Sept. 28 that SpaceX intends to sunset Falcon and Dragon as it concentrates on Starship. Isaacman’s statement came from NASA leadership and matched the company’s described direction of concentrating on Starship. Gwynne Shotwell, SpaceX’s president, said in a March interview with TIME that the company forecasts about 140 to 145 Falcon flights for this year, below last year’s 165.

Shotwell’s March forecast to TIME showed Falcon 9 is still flying at a high cadence even as the wind-down approaches.

Falcon 9’s annual flight count reached 60 in 2022, 91 in 2023, 132 in 2024 and 165 in 2025 as the rocket’s cadence accelerated. The standard launch price increased 19 percent over five years to reach $74 million. SpaceX intends to build up a stock of Falcon rockets before halting production, while government missions may extend flights into the early 2030s. The stockpile plan suggests Falcon 9 may not disappear immediately for all customers even as commercial availability tightens.

SpaceX had stopped taking commercial rideshare bookings past 2028 (one rocket carrying payloads from multiple customers), according to a July 20 SatNews report that was one of the first public indications of the restriction. Bloomberg reported four days later that SpaceX was not taking future Falcon 9 rideshare reservations and was rejecting certain dedicated orders scheduled after 2028. Ron Epstein said Falcon 9 rideshare bookings are nearly exhausted through the end of the decade, and he addressed future availability during the morning keynote at the conference.

Portal Space Systems bought one complete Falcon 9 mission for 2028 after CEO Jeff Thornburg said the company could not risk losing access to orbit, a move it had not initially planned because it could not afford a gap in access. Starship is expected to take over much of Falcon 9’s role, leaving commercial customers to compete for launches against SpaceX’s Starlink satellites and orbital computing infrastructure plans.

Over the last three years, Earth saw an average of 270 orbital rocket launches annually, a more than threefold increase compared with a decade ago. Caleb Henry, director of research at Quilty Space, described the industry as being in panic on Aug. 17, and he said satellite operators are describing an industry in panic while he speaks with them constantly about launch availability. Quilty Space has charted a shift in the discussion from how many launch companies the U.S. market can support to who can execute fastest.

The launch crunch may worsen over the next two to four years as demand continues to grow while supply remains constrained by vehicle transitions. Rocket Lab’s Neutron cannot replace Falcon 9’s capacity overnight, even though the vehicle is positioned as an opening created by the wind-down.

Telesat received inquiries about sharing some of the 11 Falcon 9 launches it booked for a new constellation, but declined, a reply that illustrates scarcity even for operators with existing bookings. Amazon’s LEO constellation has slowed satellite production because most of the dozens of launches booked five years ago are not yet ready, with production delays linked to launch readiness rather than manufacturing alone. AST SpaceMobile officials said during an earnings call last Monday that launch availability is the pacing item for deploying its constellation, a constraint affecting deployment schedules for operators.

The Commercial Space Federation forecasts thousands of satellite launches annually in less than a decade, while Analysys Mason projects more than 37,000 satellites need launching from 2023 to 2033. Ron Epstein, managing director at Bank of America Securities, said launch prices have begun rising again due to a shortage of available launches. Mike Collett, managing partner at Promus Ventures, said he finds it difficult to imagine launch costs not being higher in a year.

Epstein said Bank of America’s bull case for SpaceX already assumes Falcon 9 goes away, pushing customers onto Starship, but Starship capacity for other industry players remains an open question. Epstein made the point about BofA’s bull case during the morning keynote at Silicon Valley Space Week. Collett made the remark during the panel “After SPCX: The Public Market’s Verdict on Space,” moderated by Hogan Lovells partner Alexis Sáinz, and Promus Ventures has invested in Rocket Lab.

Dr. Paul Struhsaker, CEO of Arrasar Partners, said orbital data centers will not fly unless launch costs fall. “If launch costs do not fall, then orbital data centers will not fly.” Struhsaker made the point in a technical brief before the orbital data center panel at the conference. Cowboy Space COO Joseph Yaffe said ride-along availability is vanishing quickly, speaking on the afternoon’s orbital data center panel at the conference after Cowboy unveiled its Big Iron rocket the night before.

Epstein said there is likely a strong market for another company capable of successful launches, and he named Rocket Lab’s Neutron as a vehicle to watch, seeing an opening in the launch squeeze during the conference. Collett cautioned that few launch companies exist and that building a new one requires years and capital, a more cautious view than Epstein’s on how fast new supply can arrive.

Rocket Lab’s Neutron rocket can lift roughly 13 metric tons to low Earth orbit, and the company has a backlog of more than 90 missions, including five dedicated Neutron flights for one customer. Rocket Lab is pursuing an $8 billion deal to acquire Iridium Communications Inc., adding a satellite network with 2.6 million subscribers, and Iridium shareholders are scheduled to vote on the deal Thursday; merger filings indicate Iridium would become a customer for Neutron, giving the vehicle a guaranteed stream of launches. Rocket Lab is following SpaceX’s playbook beyond launch. Polymarket traders currently estimate a 9 percent chance that Neutron will launch by Dec. 31, based on $11,000 in trading volume.

Several companies, including Blue Origin, Relativity Space, Stoke Space and Firefly Aerospace, are developing new launch capacity, but industry executives told the Wall Street Journal that meaningful alternatives remain years away. The Journal reported the wind-down and the competitive landscape. The Pentagon’s Office of Strategic Capital has conditionally committed a $150 million loan guarantee to fund a second Stoke Space launch pad in Florida this week. Epstein said he wants to see whether SpaceX catches both stages on the next Starship flight, a step toward reusability that the transition depends on.

Canada lacks a medium-lift rocket ready to fill the gap created by SpaceX ending Falcon 9 bookings after 2028. Telesat’s Lightspeed constellation, connected to a $2.3 billion CAD federal contract covering an Arctic military layer, needs 225 satellites and has booked launches on SpaceX before the Falcon phaseout. Launch is expected to consume roughly 20 percent of the Telesat program budget, or about US$1 billion.

Prime Minister Mark Carney warned Canada should not rely on one foreign provider for space access, while U.S. Ambassador Pete Hoekstra said Canada chose to close its market to SpaceX; both officials spoke at the same Gatineau conference last week. Ottawa is investing about $200 million CAD in the Canso, Nova Scotia, launch pad operated by Maritime Launch Services under a 10-year contract, while a separate competition is dividing $25 million CAD among three small-rocket startups.

Canada Rocket Company, an Etobicoke startup, has raised $22.5 million CAD without reaching orbit, and CEO Hugh Kolias said developing a new rocket generally requires six to eight years and about $1 billion CAD to reach orbit, so the gap will not close by 2028. SpaceX recently completed a static fire of a booster with all 33 engines before Starship Flight 15, showing the company keeps pushing Starship forward while Canadian alternatives remain grounded.

SpaceX will launch Falcon 9 from Vandenberg Space Force Base between 9:45 p.m. PT Monday, Oct. 26 and 1:45 a.m. Tuesday, Oct. 27, marking the 72nd California launch of 2026 and surpassing 2025’s launches. The mission will deploy 27 more Starlink satellites, and the booster is making its 26th flight from SLC-4E at Vandenberg Space Force Base.