It's Time To Admit It: Starship Is An Embarrassing Failure
It's not just the failed test launches; Starship simply isn't what was promised.

On January 16, the residents of Turks and Caicos were treated to a display of truly biblical-looking fireworks as a SpaceX Starship disintegrated above their heads during its seventh test launch. In the past, I and many others have stated that we shouldn’t be alarmed by failed rocket experiments, given that they are a standard and expected stage of rocket development. This is still somewhat true, but it is quite concerning that a launch that fails so spectacularly–to the extent of not even reaching orbit–occurs this late into the development process. At this point, SpaceX should be optimising Starship, not desperately trying to keep it alive. On top of all of that, there is growing evidence to suggest that Starship can’t come close to its proposed specifications. In fact, SpaceX’s competition is inches away from overtaking them, and Starship is actually worse than 60-year-old rocket technology. Don’t believe me? Well, let’s take a look.
Did you know that Starship has a serious engine problem?
Incredibly few people seem to know that Musk has publicly stated that, due to SpaceX’s current issues with its engines, Starship can only take “40–50 tons to orbit.” That is less than half of what was promised! Because of this shortfall in performance, Starship would be incapable of conducting any of its proposed missions. To give a sense of scale here, SpaceX’s own Falcon Heavy rocket has already carried 57 tonnes to low Earth orbit for a fraction of the cost of Starship.
Musk has unsurprisingly failed to disclose the actual reason for this failure, other than citing that the engines are not producing enough thrust. To many, this is not surprising; the thrust figure Musk claimed for the Raptor engines used by Starship seemed wildly optimistic (bordering on fanciful). This is especially significant when you factor in that these are not single-use engines but ones expected to launch and land Starship repeatedly. Just like how your car wouldn’t last very long if you drove everywhere at full throttle, rocket engines need to be held back to extend their life. But these engines have also been repeatedly failing during these test launches, so if SpaceX has reduced their thrust to increase reliability, it isn’t even enough to make them reliable enough for single use.
It doesn’t matter if this lack of thrust comes from a fundamental issue with the rocket design, bad thrust predictions, or reliability issues. Fixing it is going to take a lot of time and cash–if it can even be fixed at all.
And this isn’t the only problem plaguing Starship. The rocket is also significantly more expensive than Musk is letting on.
Musk has repeatedly claimed that the nominal cost of building and launching a fully stacked Starship (that is, a Starship with its booster) is $100 million. There are plenty of reasons to believe this is a flat-out lie, not only because of Musk’s track record of fudging the truth but also the simple economy of building such large structures. However, the best evidence we have of the cost is SpaceX’s annual expenditure for Starship, which was $5 billion last year. If we generously assume the Starship project has overheads of a billion dollars a year and yearly R&D costs of two billion dollars, as SpaceX has suggested, we can estimate that SpaceX spent $2 billion on Starship launches last year. There were four Starship test launches last year, meaning it likely costs more like $500 million per test launch!
Moreover, only one of the test launches in 2024 reached orbital velocity, and it had zero payload! That’s right; these Starship launches only carried themselves into space and absolutely nothing else.
With this reduced theoretical payload to LEO and our more realistic launch cost of Starship, we can compare its current progress to other rocket systems.
Using Starship’s current theoretical payload to LEO (low Earth orbit) and our estimation of its actual launch cost, we can calculate that Starship costs around $10,000 per kg delivered to LEO. By comparison, the Saturn V rocket used by the Apollo program costs $9,219 per kg to LEO, adjusted to today’s currency. In other words, Starship is likely as expensive, if not more costly, than a rocket NASA developed and used before pocket calculators were invented. SpaceX’s own Falcon Heavy can carry more to LEO than the current Starship, yet it only costs $1,700 per kg to LEO, or less than 20% of the cost of Starship!
However, the most embarrassing comparison is that of Blue Origin’s partially reusable New Glenn, which can carry the same payload to LEO as Starship and is predicted by third parties to cost only $1,500 per kg to LEO! Blue Origin successfully reached orbit on its first test flight of New Glenn but failed to land its booster, meaning that the company is still a few hurdles away from achieving that price but is almost certainly closer to that predicted price than Starship is. Why? Because landing and recovering the booster (also known as the first stage) is far easier than landing the second stage (also known as the orbital vehicle).
This is why when people say Starship will get substantially cheaper as soon as we can reuse the entirety of the rocket, we should take their analysis with a pinch of salt because this is the other major issue with Starship. It likely can’t be fully reusable.
If you don’t know why rockets have multiple stages, here is a good explanation video.
Landing the first stage, like SpaceX and Blue Origin do, is relatively easy. In fact, NASA developed and tested the technology in the ’90s. This is because the booster travels significantly slower than the rocket’s second stage, which accelerates further after they have separated, meaning it has considerably less kinetic energy to disperse, making it far, far easier to land.
Starship’s booster is travelling at 5,370 mph at the point of separation, where it turns around and lands back on the ground. Meanwhile, Starship’s reentry speed is over 17,000 mph, and it will theoretically weigh more than the booster did at the point of separation. This means that Starship has exponentially more kinetic energy to manage than the booster if it wants to land safely with zero speed.
Starship is designed to re-enter belly-down and use atmospheric resistance to disperse this energy and slow down to the point where retrorockets can take over and conduct a controlled landing. However, this causes significant heating to the Starship, and thanks to its weight and immense speed, there is enough kinetic energy during reentry to literally melt the whole of the Starship several times over through atmospheric heating.
This is why Starship, like the shuttle, uses a ceramic heat shield to protect the vehicle from this heat during reentry. Unfortunately, Starship famously has a problem with its heat shield, as the tiles it is made from keep falling off, likely due to thermal contraction during takeoff and exposure to space, rendering it inefficient and exposing the rocket to dangerous heat levels during reentry. And, even if the tiles stayed on, many engineers have pointed out that with the amount of kinetic energy that Starship has to disperse, no heat shield would be effective enough to protect it or ensure a safe landing.
That is why many have speculated that the reason we don’t see a video feed from the inside of Starship during its reentry is because the inside would glow red with heat! These same people have also speculated that this is why SpaceX is still doing “splashdown” landings in the remote Indian Ocean and publishing the landing velocity of the experiments, as all that is needed to quantify that statistic is to not totally disintegrate during reentry, which even the single-use Saturn V rockets were able to achieve. However, if Starship attempted a proper landing, it would expose just how horrifically underperforming and dangerous the landing is.
So, evidence is mounting that we shouldn’t expect Starship itself to be able to land and be reused any time soon, let alone safely take humans to space and back again. Especially when you consider it is currently failing to come anywhere close to that expectation while carrying zero payload.
But the booster, or first stage, can be reused, right? That will make Starship cheaper than it currently is.
Well, yes. Ish.
There is a huge question of reliability, particularly as the engines are failing to produce enough thrust, and multiple engines are failing mid-flight, as we have already discussed. SpaceX did reuse a booster for the latest test and managed to land it, so we know they can possibly be reused a handful of times, but their long-term reusability is still massively up for debate. Moreover, SpaceX has said that the booster design will soon have to change dramatically, so we know they won’t be reused much over the coming years anyway.
As such, Starship’s overall reusability is still profoundly questionable, and its launch costs will likely be very high for a long time to come. This means Starship right now is no cheaper or more capable than the Falcon Heavy and has a long, long way to go before its price and payload can come anywhere close to what was promised (if it ever can), let alone conduct human space flight.
Ultimately, even after seven launch tests, Starship is currently a very expensive and dangerous way to take zero payload to a uselessly low orbit. On top of that, it seems highly likely to never reach its proposed payload capacity, levels of reusability, or price point. Meanwhile, NASA’s SLS successfully delivered a payload to orbit the moon on its first test flight.
It’s time to admit it: Starship is looking like one hell of a disaster. There may still be hope for this project, but it increasingly looks like the entire concept was fundamentally flawed.
Thanks for reading! Content like this doesn’t happen without your support. So, if you want to see more like this, don’t forget to Subscribe and help get the word out by hitting the share button below.
Sources: New Scientist, BBC, The Telegraph, Americaspace, Space