There is a number that haunts every conversation about the future of spaceflight, and it is not a velocity or an altitude. It is a price: roughly $10,000 per kilogram to low Earth orbit, which is approximately what it cost to launch mass into space for most of the post-Apollo era. That number is the reason space telescopes are engineered within an inch of their lives and then launched once, never to be repaired if something goes wrong. It is why satellite constellations were a fantasy for most of the twentieth century, why Mars has remained a destination rather than a waypoint, and why the economics of everything beyond the atmosphere — communications, Earth observation, scientific research, eventually human settlement — have been shaped by a cost curve that barely moved for fifty years.
SpaceX spent the last several years trying to break that number. With the completion of Starship's first fully successful orbital test flight, including the recovery of both the Super Heavy booster and the Starship upper stage, the company has made a credible case that the number can be reduced by an order of magnitude or more. Whether that reduction materializes on the timelines SpaceX projects is genuinely uncertain. That it is now plausible, mechanically demonstrated, is not.
What the vehicle actually is
Starship is not a rocket in the conventional sense — or rather, it is a rocket in the way that a cargo ship is a boat. The system consists of two fully reusable stainless-steel stages. The first stage, Super Heavy, stands roughly 71 meters tall and is powered by 33 Raptor engines burning liquid methane and liquid oxygen, producing approximately 74 meganewtons of thrust at liftoff — nearly double the Saturn V's 35 meganewtons. The upper stage, also called Starship, adds another 50 meters and six Raptor engines of its own. Together, the stack clears 120 meters and represents the highest mass-to-orbit capacity ever attempted for a fully reusable system.
The Raptor engine itself deserves attention. It runs on a full-flow staged combustion cycle, a thermodynamic approach that extracts more energy from propellant than the open-cycle engines that powered most twentieth-century rockets. Soviet engineers attempted full-flow staged combustion in the 1960s and found it extraordinarily difficult to execute reliably. SpaceX solved it, scaled it to production, and then designed a vehicle that requires 33 of them to fire simultaneously and remain stable through max dynamic pressure. The early test flights, in which boosters exploded on the pad or disintegrated shortly after launch, were not failures of ambition — they were an unusually public version of the iterative development process that any complex propulsion system requires.
Recovery changes the arithmetic
The economic case for Starship rests almost entirely on reusability. A Falcon 9, SpaceX's current workhorse, costs somewhere in the range of $67 million per launch and can carry about 22.8 metric tons to low Earth orbit in its expendable configuration — less with booster recovery factored in. SpaceX has driven that cost down dramatically from the industry average through vertical integration and booster reuse, but the upper stage is still discarded on every flight, which sets a floor on how cheap the system can become.
Starship's design eliminates that floor. If both stages can be refurbished and reflown — and SpaceX has now demonstrated recovery of both in the same mission — the marginal cost of a launch becomes dominated by propellant, refurbishment labor, and the amortized cost of the vehicle across its flight life. Elon Musk has stated target launch costs below $10 million, and while that figure should be treated with appropriate skepticism given SpaceX's history of optimistic timelines, the underlying engineering logic is sound. Propellant for a full Starship mission costs on the order of $900,000. The rest is operations.
NASA has already internalized this logic. The agency awarded SpaceX a Human Landing System contract worth up to $4 billion specifically because Starship's cargo capacity — nominally 100-150 metric tons to low Earth orbit in a fully reusable configuration — makes the Artemis lunar architecture viable in a way it would not be with existing vehicles. The Lunar Starship variant foregoes reusability for maximum performance in the cislunar environment, but the basic premise holds: if Starship works at scale, the cost per kilogram to the Moon's surface could fall to a level that makes sustained human presence there economically discussable rather than merely aspirational.
The industry reorganizes around a possibility
The more immediate effects are visible in how the rest of the launch industry is positioning itself. United Launch Alliance's Vulcan Centaur, which made its certification flights in 2024, is a capable vehicle, but it is only partially reusable — the BE-4 engines are recovered, the rest of the first stage is not. Blue Origin's New Glenn achieved orbit in early 2025 with a reusable first stage, a genuine milestone, but its upper stage is expendable and its payload capacity is well below Starship's even in optimistic projections. Ariane 6, Europe's answer to the post-Falcon-9 market, is largely expendable and has struggled with its own development delays.
None of these vehicles are obsolete — they serve real missions and real customers who have reasons to prefer them, including contractual relationships, security considerations, and the straightforward desire not to depend on a single launch provider. But the competitive dynamic has shifted. If Starship achieves routine operations at something approaching its projected economics, every other vehicle in the market faces pressure to justify its per-kilogram cost against a standard that did not exist five years ago. Some of that pressure will accelerate development of next-generation reusable systems. Some of it will simply change what kinds of missions get funded and flown.
The satellite industry is the most immediate example. Starship's cargo capacity means that spacecraft no longer need to be engineered for minimum mass as a primary constraint. Instruments can be larger, redundant systems can be included, and constellations can be deployed in configurations that would have been prohibitively expensive to launch under the old cost structure. For scientific missions, this is not a minor convenience — it is a potential shift in what questions can be asked. A space telescope with a primary mirror several meters across costs roughly as much to design and build whether it is launched on a vehicle that can afford the mass budget or not. The launch cost is what determines whether it gets funded.
There is also the question of orbital refueling, which Starship's architecture requires for deep-space missions and which has never been demonstrated at operational scale. NASA's Artemis mission profile calls for multiple Starship tanker flights to fill the lunar variant before it departs for the Moon. This is a significant technical challenge — transferring cryogenic propellant in microgravity involves managing fluid dynamics that behave differently than they do on the ground — and it remains one of the genuinely open engineering questions in the program. SpaceX has conducted early propellant transfer demonstrations, but the full operational sequence for a lunar mission has not been tested end-to-end.
That caveat matters, but it should not obscure what the orbital test flight represents. Rocketry is an industry in which the gap between a successful demonstration and routine operations is often measured in years and billions of dollars. The history is full of vehicles that flew successfully and never achieved their economic promise. Starship may be another one. But it is now in the category of things that have been shown to work in the physical world, which is a different category than things that are projected to work in a slide deck. The next several flights will establish whether the refurbishment cycle is as fast and cheap as SpaceX needs it to be. The answer to that question will determine whether the $10,000-per-kilogram wall is cracked or demolished.
