Stories & guides · · 27UsedCar

EV Battery Health Explained: What State of Health (SoH) Really Tells You

SoH is the most useful single number for judging a used EV's battery. Real-world data from tens of thousands of cars shows how fast batteries degrade, what large samples observed around 90,000 km, and which factors those studies link to faster ageing.

Electric vehicle battery health

State of Health, or SoH, is the single most useful number for judging a used electric car's battery. This guide explains what SoH means, how fast batteries actually degrade according to large real-world datasets, what speeds up ageing, and how to use SoH when comparing used EVs.

What SoH means

Every battery starts life at 100% state of health. Degradation is the natural, permanent loss of the energy a battery can store (and, less visibly, the power it can deliver). A battery's condition is expressed as SoH: a 60 kWh pack at 90% SoH effectively behaves like a 54 kWh pack, and at 80% SoH like a 48 kWh pack. In practice, SoH loss shows up mainly as reduced range, and on some models as slower DC fast charging once degradation becomes advanced.

How fast do batteries really degrade?

Large fleet datasets suggest modern EV batteries age more slowly than many buyers fear. Geotab's updated analysis of more than 22,700 electric vehicles across 21 models found an average degradation rate of about 2.3% per year — slightly higher than the 1.8% per year in its earlier study, which Geotab attributes in part to newer vehicles using higher-power charging. A separate 2026 study by TÜV Nord and Carly, based on around 50,000 SoH readings from cars built between 2016 and 2026, found a median SoH of 96 points, with only about one in ten vehicles below 85 points — the level beyond which losses in range and charging behaviour become more noticeable.

What one large sample observed around 90,000 km

In the TÜV Nord and Carly sample, the average loss was about 0.7 SoH points per 10,000 km up to roughly 90,000 km, and about 2.3 points per 10,000 km beyond that mileage. That is an observation about that sample — a mix of models, ages and usage patterns measured with one provider's method — not a universal rule that every battery changes behaviour at 90,000 km. Degradation curves differ by model, battery chemistry, climate and how SoH itself is measured, so treat the figure as context for why a documented reading matters, not as a pass/fail line for an individual car.

What speeds up battery ageing

In Geotab's fleet sample, charging behaviour was the strongest operational factor associated with degradation. Vehicles that relied heavily on DC fast charging above 100 kW degraded at up to about 3.0% per year on average, compared with roughly 1.5% per year for vehicles charged mainly on AC or lower-power chargers. Climate played a smaller role in the same dataset: cars in hotter regions degraded about 0.4% per year faster than those in mild climates. Using a wide state-of-charge range in daily driving was not, by itself, associated with meaningfully higher degradation; prolonged, habitual periods near completely full or completely empty were. These are averages across a fleet sample — they describe groups of vehicles, and they cannot be run backwards to diagnose how one individual car was charged by its previous owner.

Thermal management design matters too. Geotab's earlier model comparisons showed liquid-cooled packs ageing markedly more slowly than passively air-cooled ones — its example compared a liquid-cooled 2015 Tesla Model S (about 2.3% per year) with an air-cooled 2015 Nissan Leaf (about 4.2% per year).

How to use SoH when buying a used EV

Ask for a measured SoH value from a proper diagnostic read, not a guess from the dashboard range estimate, which fluctuates with weather, driving style and software settings. Before comparing numbers, establish what the number means for that car: manufacturers and test providers do not all calculate SoH the same way — some reference total capacity, others usable capacity — and test temperature, recent driving and software version can all move a reading. The meaningful comparisons are therefore like-for-like: the same model, tested the same way.

Then judge the reading against the manufacturer's own standard for that model, not against a universal threshold. The relevant document is the car's battery warranty, which states the capacity level and the age/mileage limits the maker guarantees; a result has a different meaning on a car still inside that warranty than on one outside it. If the seller cannot provide a measured reading, a professional battery test before purchase is the practical substitute. Finally, keep SoH in context: charging speed on a real charger, warranty status, and service history complete the picture, and no single percentage prices a car on its own.

Sources

Sources

Editorial review: 27UsedCar ·

A clearer view of the market.

Used cars, shipping updates and buying guides, in your language.

By subscribing, you agree to receive market updates. Unsubscribe any time. Privacy Policy

WhatsApp