EV battery degradation data is finally good enough to price a used EV accurately

For most of the last decade, pricing a used electric vehicle meant pricing an unknown. A used gas car's engine wear is roughly legible from mileage and maintenance records; a used EV's most expensive component — the battery pack — degrades on a curve that depends on chemistry, climate, charging habits, and age in ways no odometer reading captures. Dealers defaulted to steep, blanket depreciation because they had no better model. That's changed. Millions of real-world packs now have years of telemetry behind them, companies like Recurrent and major fleet operators have published aggregate degradation data at scale, and the result is a degradation curve accurate enough to actually price a specific used EV rather than guess at its battery's remaining life.

What the aggregate data actually shows

The headline finding across large fleet studies is that EV battery degradation is front-loaded and then flattens. A pack typically loses a few percent of usable capacity in its first year or two — part of this is genuine chemical break-in, part is simply that early capacity estimates from the car's battery management system are conservative — and then the degradation rate slows to roughly 1-2 percent of capacity per year for the remainder of a well-managed pack's life. That means a 5-year-old EV with 60,000 miles on an LFP or well-thermal-managed NMC pack is very often still sitting at 88-92 percent of its original range, not the 70 percent that older, worst-case assumptions implied.

Chemistry is the biggest single variable the data separates out. LFP (lithium iron phosphate) packs — used by Tesla's standard-range models, most Chinese EVs, and increasingly Western automakers chasing lower cost — degrade more slowly and tolerate frequent fast-charging and full charges better than NMC (nickel manganese cobalt) packs, which have historically offered higher energy density at the cost of being more sensitive to heat and to being kept at a high state of charge for long periods. The practical upshot: an LFP-pack EV with heavy fast-charging history is a safer used buy than an equivalent-mileage NMC EV with the same fast-charging pattern.

Climate and thermal management explain most of the variance that chemistry alone doesn't. Packs without active liquid thermal management — common on several early and budget EV models — show measurably faster degradation in markets with extreme summer heat, because heat accelerates the same chemical side-reactions that cause capacity loss over time. A used EV's climate history (was it driven and charged somewhere that regularly exceeded 35°C) is now a more useful input to its expected remaining range than its total mileage.

Why this data wasn't available until recently

The lag was structural, not a failure of research. Meaningful degradation curves require years of real-world data across thousands of vehicles, charging patterns, and climates — you can't shortcut that with lab testing alone, because lab cycling tests don't replicate the irregular, real-world charge patterns (partial charges, fast-charging frequency, idle time at high state of charge) that actually drive degradation. The first wave of mainstream EVs with meaningful fleet size — Model S/3, Leaf, Bolt — only crossed the 5-8 year mark where degradation curves become statistically legible in the past few years, and telematics access to battery state-of-health data at scale is a relatively recent capability for third parties who aren't the manufacturer.

How to actually use this when buying or selling a used EV

• Pull the actual state-of-health number, not just mileage. Tools like Recurrent's battery reports, or the car's own app/diagnostic menu on many models, report a current usable-capacity estimate. Treat this as the primary pricing input and mileage as secondary.
• Weight chemistry into your risk tolerance. An LFP pack with 70,000 miles of mixed charging is a lower-risk buy than an NMC pack with the same mileage and a history of frequent fast-charging to 100 percent.
• Ask about climate history, not just accident history. A CarFax-style report tells you about collisions; it tells you nothing about whether the car spent five summers in Phoenix with no shade and no active thermal management. If you can get charging-location history, it's now a more predictive variable than you'd expect.
• Expect the front-loaded dip, don't panic about it. If a 1-year-old EV shows 94 percent instead of 100 percent capacity, that's consistent with normal early degradation and conservative initial BMS estimates — it's not evidence of a bad pack, and it shouldn't move the price much.
• Discount heavily for packs with no active thermal management in hot climates with high mileage — that's the single combination the fleet data flags as the real risk case, more than age or mileage alone.

The practical takeaway is that "used EV" has stopped being a single risk category. The data now supports pricing a specific car based on its chemistry, its climate history, and a measured state-of-health number — and buyers or sellers who ignore that and price purely on mileage are leaving real money on the table in either direction.