If you are shopping for a used electric car, someone will eventually show you a battery percentage and expect it to settle the question.
It does not, and there are two separate reasons why. The first is that people routinely mix up two different numbers. The second is that the number they usually mean is an estimate rather than a measurement.
Neither of those is a scandal. Both are worth understanding before you spend thirty thousand dollars.
State of Charge and State of Health are different things
State of Charge is how full the battery is right now. It is the fuel gauge. It reads 80% after a charge and 20% before you plug in, and it tells you nothing about the condition of the battery.
State of Health is how much total capacity the pack still has compared to when it was new. A battery at 90% State of Health holds 90% of its original capacity. That is the number that matters when you are buying, and it is the number people mean when they say "battery health."
You can have a completely worn out battery showing 100% State of Charge. It is full. It is just that full does not mean what it used to.
When a seller sends you a screenshot, the first thing worth checking is which of these two numbers you are actually looking at.
The health number is an estimate, not a measurement
This is the part that surprises people, and it is not a criticism of any manufacturer.
Your car does not directly measure the condition of its battery cells. It cannot. Instead, the battery management system calculates an estimate, using models built from laboratory data along with what it has observed about voltage, temperature, current and how the car has been used.
That estimate is generally reasonable. It is also a calculation rather than a reading, produced by the vehicle's own software, and it can drift.
So the honest way to hold that number is as a well-informed opinion from the car about itself. Useful. Not the same as an independent result.
What normal degradation actually looks like
Large fleet studies give us a decent picture here.
Geotab analyzed more than 22,700 electric vehicles and found average degradation of roughly 2% per year, which projects to about 81.6% state of health by year eight. Recurrent, working from over 1.6 million vehicle observations, lands in a similar range at 1 to 2% per year.
So as a rough sanity check:
| Age | Roughly what to expect | |-----|------------------------| | 2 years | around 96% | | 4 years | around 91 to 96% | | 6 years | around 88 to 92% | | 8 years | around 82 to 88% |
Individual cars vary a lot. Climate, fast-charging habits, and how the car is stored all move the number, and averages across tens of thousands of vehicles will never predict one specific car.
Use this as a smell test, not a verdict. A four-year-old EV reading in the low nineties is entirely normal and nothing to worry about. A four-year-old EV reading a flat 100% is unusual against fleet averages, which does not prove anything is wrong, but it is a reason to look more closely rather than take the figure at face value.
If you are looking at a used Tesla
Tesla has a battery health test built into the app, which is genuinely useful and more than most manufacturers offer.
Two practical limits worth knowing. It is rate limited, generally to about once every six months outside of service mode, so a seller cannot simply run it on request while you stand there. And running it puts the car through a full charge cycle, which takes real time and is not something most sellers will do on a Saturday afternoon for a stranger.
None of that makes the result untrustworthy. It just means the test is often not available at the moment you actually need it.
Why a seller's number is a hard thing to rely on
Put the technical detail aside for a second, because the real problem is simpler.
Even if the number is completely accurate, you have no way to verify it and no way to prove it afterwards. You are being asked to accept the most expensive component in the car on the word of the person selling it to you.
That is not an accusation. It is just how any private sale works, and it is the same reason people get a home inspection rather than trusting the seller's description of the roof.
What independent testing adds
An independent test produces something the dashboard cannot: a document from a party with no stake in the sale.
There is a common assumption that the point of an independent test is to catch the car out. Usually it is not. Good independent tests tend to agree with the vehicle's own figure reasonably closely, and that is a sign the test is working rather than a sign it was pointless.
The value is not disagreement. It is confirmation you can actually rely on, written down, that travels with the car and can be handed to a buyer, a seller or a lender.
What to do before you buy a used EV
- Check which number you are being shown. Charge or health. They are not the same.
- Compare it against the car's age. Roughly 2% a year is the fleet average.
- Be curious about a perfect score on a car that is several years old. Not alarmed. Curious.
- Ask when the figure was produced, since a screenshot from months ago describes a different car than the one in front of you.
- Get it verified independently if the purchase is large enough that being wrong would hurt.
We run independent EV battery health tests as part of our pre-purchase inspections in Atlanta, using AVILOO, whose testing is certified by both TUV Nord and TUV Sud. We can test Tesla and 57 other EV brands. A few, including Rivian, Lucid and the Tesla Cybertruck, are not covered yet, so it is worth checking your vehicle before booking.
Our technicians are certified and factory trained, and we come to wherever the car is.
Book an inspection or call 833-644-0800.