A used electric car hands you one number to trust, and a British battery-diagnostics firm has just published evidence that the number leaves something out. Generational analysed roughly 10,000 used-EV battery tests taken through hundreds of UK dealers and looked past State of Health at the individual cells inside each pack. Most were fine. A small, consistent minority were not, in a way State of Health does not report.
What the cell readings showed
A healthy pack keeps its cells at nearly the same voltage. Across the sample the median gap between the highest and lowest cell was eight millivolts — effectively nothing.
| Cell voltage spread | Share of tests | What it indicates |
|---|---|---|
| ~8 mV (median) | typical | Normal, well-balanced pack |
| Above 50 mV | 1 in 85 | Significant imbalance worth investigating |
| Above 100 mV | 1 in 200 | Serious imbalance, likely an underlying fault |
Above about 100 mV the consequences are the ones an owner actually notices. Range becomes less predictable, charging can misbehave, and the battery management system starts capping performance to protect the cells that are drifting. Nothing on the dashboard says "cell imbalance"; the car simply does less than it used to.
Auto Express scaled the one-in-85 figure against the two million-plus EVs on British roads and arrived at more than 25,000 affected cars. That is arithmetic on a sample rather than a count of faults found, and worth reading as an order of magnitude rather than a census.
Why State of Health misses it
The two measurements answer different questions. "Reporting on State of Health is essential for used-EVs, but it does not tell the full story of battery condition," said Oliver Phillpott, Generational's chief executive. "It measures capacity, not necessarily whether the cells inside the pack are functioning correctly."
That distinction matters because State of Health is the metric the whole used market has standardised on. China now has a national battery-health yardstick and Europe's equivalent arrives in 2027 — both of them measuring capacity. A pack can report 93% State of Health, look entirely respectable on a dealer's printout, and still contain a cell drifting 120 mV away from its neighbours.
The firm's suggested answer is unglamorous: test annually, so drift is caught while the battery is still under warranty rather than after it. Generational offers a guided rebalance for packs it flags, and WhichEV reports that roughly half of the imbalances identified resolved after a controlled rebalance, with the rest needing deeper investigation or replacement.
What a used-Tesla owner in Europe should take from this
Tesla is not named in the analysis, but it is unavoidably in the sample. Britain bought a record 110,761 used electric cars in the second quarter of 2026, and Tesla takes three of the ten fastest-selling used EV models in the country. The Model 3 and Model Y are the volume used EVs on British forecourts, so they are the cars this finding is most likely to be describing.
Tesla packs have a good record on the capacity question — a Swedish study found the LFP Model 3 battery held up best after 100,000 km. Cell balance is a separate axis, and one that capacity testing was never designed to expose.
The practical takeaway for a used-Tesla buyer is narrow but real: a good State of Health reading is necessary and not sufficient. If a seller can produce a cell-level voltage report alongside it, that is a materially better disclosure than the single percentage almost everyone is currently working from — and on a one-in-85 base rate, worth asking for.