In 2022 NASA conducted a successful uncrewed lunar flight of the Orion, Artemis 1. Yet, the following mission and first crewed flight, Artemis 2, will launch in fall 2025 the earliest. In comparison, SpaceX's Starship had two launches within about half a year (and it took that long only because of the FAA's investigation of what to improve), the first one was brought to destruction after insufficient thrust and the second one too after telemetry was lost, and it plans to conduct the third test flight as soon as the FAA allows it again.

SpaceX had to build their BFR/Starship anew, as the first one was destroyed during the test flight and the second Starship was destroyed as well. And the BFR/Starship is bigger than the SLS, isn't it? So why isn't NASA able to conduct flights as frequent as SpaceX, especially since Artemis 1 was a full success?

  • $\begingroup$ "SpaceX had to build their BFR/Starship anew, as the first one was destroyed during the test flight and the second Starship was destroyed as well." They are not planning to recover any of the early test rockets, so building new Starships for each test flight is part of the plan. "it took that long only because of the FAA's investigation of what to improve" that is a supposition and there is a lot of evidence to the contrary. However your overall question is valid (why Starship flies more often than SLS). $\endgroup$ Commented Feb 22 at 14:31
  • $\begingroup$ @StevePemberton What I wanted to say is that despite it's not reused (so far), SpaceX can afford to build everything anew within a few months and launch it again. As for the 2nd quote, I only mean that the launch date of IFT-2 was dependent on the investigation, not on the time to build the rocket (which had already been done). $\endgroup$ Commented Feb 22 at 16:00
  • $\begingroup$ SpaceX built, not just one Starship stack, but an assembly line for producing Starship stacks (which dovetails nicely with their willingness to fly hardware to see whether it works, as well as their eventual plan to fly often). ULA was focused on producing one highly reliable vehicle at a time. (That's a an oversimplification, but they did not prioritize rapid rocket assembly the way SpaceX did.) Hopefully someone will answer who can quote reliable sources. $\endgroup$ Commented Feb 22 at 20:06
  • 2
    $\begingroup$ CharlesStaats - and maybe someone can find the quote where Elon told his staff that if the rockets are not blowing up occasionally they aren't innovating enough, or something to that effect. Of course the quote is often misunderstood since it applies only to early test designs not more mature ones, but still it's part of the philosophy, and partly explains the phase they are in now (although blowing up Starships is getting less and less desirable as time goes on). $\endgroup$ Commented Feb 22 at 20:40
  • $\begingroup$ @CharlesStaats: "SpaceX built […] but an assembly line for producing Starship stacks" – Not only that, but they are already starting to work out how to build many assembly lines. Mass-producing factories for mass-producing rockets, if you will. They're starting to "mass-produce" (for small values of "mass") launch towers. $\endgroup$ Commented Feb 24 at 15:02

4 Answers 4


NASA's goal is not to launch rockets as frequently as possible, but to achieve some wins for the American-led space program and to persevere as a well-respected and well-funded government agency. So, NASA will specify somewhat proscriptively, require extensive simulations, and demand thorough testing. What the public will see is just the end product - a series of widely-spaced (but hopefully all successful) launches that move the space program decisively forward. (additional info)

SpaceX's Starship is using more of a launch-and-learn strategy. The public is going to observe more frequent launches and, percentage-wise, more failures.

At some point, SpaceX will have a new launch system that delivers a service that people are willing to pay for. At that point, NASA will have a choice. They could continue to spend their budget on SLS launches or they could start purchasing launches from SpaceX instead.

If SpaceX can deliver more of what NASA wants with the money available in NASA's budget, then NASA will switch to using SpaceX and we will see more launches. But it is also possible that the price of SLS will come down if SpaceX enters the market with a competitive system. This could lead to more SLS launches.

For now, SLS has a solid lead. There are several challenging milestones between where Starship is today and where it needs to be to compete with SLS. Therefore, SLS's launch rate will continue to be defined by NASA's budget and its approach to designing missions and solving problems.

  • $\begingroup$ The thing is, the first Artemis flight was a full success, so why is NASA setting Artemis 2 to launch as late as late 2025? Your answer implies they're having problems with either the construction or perhaps found possible anomalies from Artemis 1 they're about to solve? $\endgroup$ Commented Feb 23 at 8:34
  • 3
    $\begingroup$ @OldManJohn Mainly risk aversion. Even though the test went okay, it was just a single test. That's hardly a guarantee everything is going to go smoothly the next time. And with actual human payload and taxpayer funding, if #2 blows up, lots of people will start questioning the whole program. It's not the same as Musk shrugging and tweeting "Nice fireworks this time, here's the money for another try". $\endgroup$
    – TooTea
    Commented Feb 23 at 10:37
  • 1
    $\begingroup$ @OldManJohn Unlike Starship, there really is no production line for Artemis vehicles. They take a long time to build. The design is optimized to keep political donors happy, not for mass production. $\endgroup$
    – John Doty
    Commented Feb 23 at 14:48
  • $\begingroup$ I was under the impression that Congress had specifically mandated the use of SLS... $\endgroup$
    – Slarty
    Commented Feb 24 at 22:30
  • $\begingroup$ That's just the government's way of ratifying the plan that NASA worked out to get everyone on board. $\endgroup$
    – phil1008
    Commented Feb 25 at 0:53

So why isn't NASA able to conduct flights as frequent as SpaceX, especially since Artemis 1 was a full success?

There are many reasons:

  • At the moment, it takes the contractors about two years to build an SLS stack. Since there is not really a production line that could accommodate lots of parts being built in parallel, this limits the flight rate to at most one flight every other year.
  • The Orion capsule for Artemis 2 uses some of the components out of the Orion capsule from Artemis 1 in order to save cost. The process of un-installing the hardware from Orion 1 and installing it on Orion 2 takes about two years. So, even if an entire fleet of SLSs had magically appeared out of thin air, there would be no Orion to put on top of it – the earliest Artemis 2 could happen would be about now.
  • For Artemis 3, the astronauts will need space suits. Development of the space suits is massively delayed, they will not be ready before the second half of 2025. So, even if a fleet of SLSs and Orions magically appeared out of thin air, NASA could have only launched Artemis 2 right after Artemis 1, then still would need to wait until late 2025 (or more realistically early 2026) for Artemis 3.
  • For Artemis 4, NASA needs SLS Block 1B with the Exploration Upper Stage, and it needs Mobile Launcher 2 to launch it. Just 4 days ago, Boeing put out a press release, according to which they haven't built a complete test article of the EUS yet, let alone a flight article. According to an internal review, the Mobile Launcher 2 is over 2.5 years delayed and \$500 million over budget already, and has only recently even started construction. According to this review, it will not be ready for launch until end of 2026. An independent external review even puts this date to late 2028. So, even if a fleet of SLS and Orion vehicles and a warehouse full of space suits appeared out of thin air, NASA could at most launch Artemis 2 and 3, then would have to wait for the EUS and the ML2 to be finished, which may be as late as end-of-year 2028.
  • There were only 16 RS-25Ds left after the Shuttle program ended, including spares and partially built ones. This means there are only enough engines for Artemis 1–4. The RS-25E is still under development. So, even if a fleet of SLS Block 1B with EUS and Orion vehicles and a warehouse full of space suits and the Mobile Launcher 2 suddenly magically appeared out of thin air, NASA could at most launch Artemis 1–4, then they would run out of engines and would have to wait for the RS-25E.
  • Last but not least: cost. Artemis is freakishly expensive, and Congress keeps cutting NASA's budget. They simply cannot afford to launch SLS that often. While Elon Musk's estimates always need to be taken with a ginormous shovel of salt, the current hope is that the launch cost for Starship could be brought down to as low as \$1 million per launch. SLS is estimated at around \$2 billion. In other words, for the price of one SLS launch, you could launch two thousand Starships!

But the most important reason is: Artemis is designed to be inefficient! NASA gets its money from Congress. With the current political climate, getting pretty much anything through Congress requires a supermajority in the Senate. Therefore, in order to get money for the Artemis mission, you need to get at least 218 Representatives and 60 Senators to agree, that means Congresspeople from at least 30 different States – most likely all from the same party, since it is unlikely you get anything bi-partisan passed at the moment, after all, the two parties cannot even agree who is the President at the moment. How do you get a Congressperson from a particular State to give money to Artemis? Mostly, by ensuring that their State gets money back, in the form of taxes, i.e., jobs in that State.

Therefore, Artemis is specifically designed to create jobs in as many different States as possible, which means produce parts in as many different States as possible. This makes production and transport very inefficient. (There is a reason SpaceX is moving from the Falcon 9 model, where vehicles are built in California, shipped to Texas for testing, then shipped to California and the Cape, to the Starship model, where every launch site gets a test site and a factory, so you build and test right next to the pads.)

Another way to get Congresspeople to vote for you, is lobbying. Only large companies can afford to spend lots of money on lobbying. Therefore, Artemis is specifically designed to use as many big contractors as possible: Boeing, Aerojet Rocketdyne, Northrop Grumman, and ULA for SLS, Lockheed Martin for the Orion Command Module, Airbus for the European Service Module, and many others.

For many different reasons, including national pride and lobbying by "old space" companies, Congress has mandated that Artemis must use SLS and SLS must be "Shuttle-derived" and re-use Shuttle hardware and designs. This already automatically excludes all the modern, agile, fast "newSpace" companies and only leaves the "old space" contractors. Therefore, Artemis is specifically designed to use as much legacy from Shuttle as possible, despite the fact it doesn't make much sense.

Artemis also uses old-style "cost-plus" contracts. "Cost-plus" means that the contractors get paid whatever their costs are, plus a fixed bonus on top. In other words: the slower they build and the more expensive they are, the more money they get. With cost-plus contracts, the contractor gets rewarded for being late and over budget.

  • $\begingroup$ That sounds as we might forget Artemis altogether. Perhaps Artemis II will occur, but the space-station-near-the-Moon plan is too ambitious. It would be better do subsidize SpaceX which already has both a mothership and a lander (as Starship is both). $\endgroup$ Commented Feb 24 at 13:38
  • $\begingroup$ That's easy to do: all you need to do is convince 54 Senators from 27 States to destroy thousands of jobs in their States and give all that money to California, Texas, and Florida. $\endgroup$ Commented Feb 24 at 13:44
  • 3
    $\begingroup$ there's also a limited number of museum-grade SSMEs to throw into the ocean forever, and Aerojet can only build SSME replicas to throw into the ocean forever so quickly $\endgroup$
    – Erin Anne
    Commented Feb 24 at 20:43
  • 1
    $\begingroup$ @ErinAnne: Thanks, added to the list. Yeah, I just read that Aerojet has proudly announced they've figured out how to shave 22 months off the build time for the combustion chamber – which implies that just building the combustion chamber alone currently takes far longer than 22 months. OTOH, SpaceX is cranking out one Raptor 2 per day. (A single engine takes longer than that to build, of course, that's the average throughput of the whole production pipeline.) Raptor 3 is supposed to simplify plumbing even more to reduce production time. $\endgroup$ Commented Feb 24 at 21:11
  • 1
    $\begingroup$ So much hangs on the wildly aspirational notion that Starship will achieve engineering miracles. If Elon said that Tesla's next EV would be 1000 times cheaper, no one would believe him. But when he says that his next rocket will be 1000 times cheaper, for some reason, some people do believe him. It makes no sense. $\endgroup$
    – phil1008
    Commented Feb 27 at 23:03

Intended mission parameters, and associated costs.

NASA never intended a rapid re-launch schedule for Artemis; even if the launcher was capable of it, the missions wouldn't be very frequent due to planning, budget, training, bureaucratic, and a number of other constraints unrelated to the launcher capacity; there was no need to design a rapid re-launch platform, and so the platform was designed without rapid re-launch capacity in mind, saving a lot of R&D costs associated with necessary streamlining and optimizing the production/preparation process.

The ultimate goal of Starship is colonization of Mars. The mission profile consists of launch of a crewed starship into LEO, then a number of 'fuel tanker' starships in rapid succession fueling the startship in orbit with fuel for the departure burn. Since they are fueling it with cryofuels which boil off over time, the shorter the refueling period the less launches are needed, but the minimum is still around 6.

That means Starship must be designed with quick turn-around in mind; the constraints go in opposite direction as NASA's; in case of NASA trying to go fast would be wasting resources, in case of SpaceX going slow would be wasting resources - and as a side effect of optimizing their process for speed, SpaceX can afford far more frequent test flights, and losing prototypes in these tests.

  • 1
    $\begingroup$ "saving a lot of R&D costs" is not something I'd associate with the SLS program $\endgroup$
    – Erin Anne
    Commented Feb 23 at 10:19
  • $\begingroup$ @TooTea Cheaper? STS was intended to have a fast turnaround, it completely failed at that goal and the costs ran higher than if it was designed with slow turnaround from moment one, as procedures and systems had to be added along the way as their necessity became apparent. You can be damn sure SLS and Artemis would be much more expensive than they are now if they were intended to have a fast turnaround, and they wouldn't necessarily be any faster anyway. $\endgroup$
    – SF.
    Commented Feb 23 at 10:38
  • 3
    $\begingroup$ @ErinAnne - ironically one of the main selling points of SLS was reduced R&D costs by reusing existing Shuttle hardware. Congresspeople, even those not in beneficial districts, fell for these simplistic arguments. I remember at the time critics were very vocal, saying that modifying the old hardware is not going to be easy, and they were proven to be right. $\endgroup$ Commented Feb 23 at 15:22

NASA doesn't want to launch another mission as quickly as spaceX does as their aim is different. NASA has to train astronauts as we've never returned to the moon or any other space body with different gravity, temperatures, etc. So NASA takes longer between their Artemis missions and SpaceX doesn't.

SpaceX has more of a 'launch as frequent as possible' aim. SpaceX wants to learn from launching more, while NASA wants to learn from the current data from Artemis1 and train astronauts

  • $\begingroup$ Artemis II is "only" a circumlunar flight, without lunar landing. SpaceX is planning a similar flight with the Starship, dearMoon. $\endgroup$ Commented Feb 25 at 6:35
  • $\begingroup$ This might be a better answer with reference to the time taken training existing Gemini astronauts to fly Apollo landing missions, Apollo earth orbit to Apollo landing and time between Apollo missions. I suspect you'll find that a 3 monthly mission cadence would work ok if budget allowed duplication (ie having multiple overlapped crews in training pipeline 'graduating' at different times) $\endgroup$ Commented Feb 25 at 7:35
  • $\begingroup$ Your answer could be improved with additional supporting information. Please edit to add further details, such as citations or documentation, so that others can confirm that your answer is correct. You can find more information on how to write good answers in the help center. $\endgroup$
    – Community Bot
    Commented Feb 25 at 8:43
  • $\begingroup$ You've posted several answers which have not included any references. A good answer here will be supported by references to show that what you write is not just your opinion or something you made up. For example, how do you know what NASA wants or what SpaceX wants? Not to mention that they are both large organizations and different parts of them probably "want" different things. If your claim is that the slow pace of Artemis launches is due to crew training, that is incorrect. $\endgroup$ Commented Feb 25 at 14:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.