There are four ways to check an EV battery's health, and they are not equivalent. The dashboard tells you almost nothing. An OBD dongle and app tells you something, if you can interpret it. An independent diagnostic test reads and interprets the battery management system's own data in minutes. A laboratory cycle test physically measures capacity and is the gold standard nobody uses for a used-car purchase.

1. The dashboard range estimate

What it measures: the car's guess at how far it will go, based on recent driving and the state of charge.

What it does not measure: capacity. A 100% charge showing 380 km on a car that once showed 450 km suggests loss, but the estimate moves with driving style, temperature, tyres and the last few trips. Two cars with identical batteries can show different numbers on the same day.

Verdict: a hint, not a measurement.

2. An OBD dongle and a phone app

What it measures: whatever the app can read from the battery management system — on some cars, a State of Health figure and cell voltages; on others, very little.

What it does not do: interpret. You get numbers. Whether 91.4% is good for a three-year-old Kona, whether a 40 mV cell spread matters, and whether that pending fault code is serious are questions the app does not answer. Coverage also varies wildly by make.

Verdict: genuinely useful for an owner who wants to watch their own car over time. A weak basis for a $30,000 purchase decision, and most sellers will not let a stranger plug an unknown dongle into their car.

3. An independent diagnostic test

What it measures: the pack's own State of Health as tracked by the battery management system, cell-group voltage spread, temperature behaviour, charging history where the car records it, and stored or pending high-voltage fault codes — read at the car, in minutes, without driving it and without writing anything to the vehicle.

What it adds over a dongle: interpretation against the model, a written report you can hold up in a negotiation, and access — a seller will let a professional plug in for a few minutes when they would not lend the car.

Its honest limit: it reads and interprets the vehicle's own management data rather than physically cycling the pack. On most modern EVs the BMS figure is reliable; on some it needs a recent full charge to be trustworthy. We set this out in full on our comparison of test methods.

Verdict: the right tool for a used-car purchase. This is what we do, for $249, at the car.

4. A laboratory cycle test

What it measures: actual capacity, by fully charging the pack and discharging it under controlled load.

Why nobody uses it for a used car: it takes the car for a day or more, requires equipment no mobile service carries, and costs far more than the question is worth on a car that is not showing symptoms.

Verdict: the gold standard, and the wrong tool for this job.

What to actually do

For a car you already own: a dongle and app, checked occasionally, will show you the trend. For a car you are about to buy: get a diagnostic read with a written report before money changes hands, and put it next to the dashboard estimate. If the two disagree, the diagnostic is the one to believe.

What a good State of Health looks like · Can you test without driving the car? · Book a test — $249


Verified 15 September 2026.