Best Cars for Resale Value in 2026: the Ranking Measures the Calendar, Not the Odometer

On 22 August 2026 I pulled the first twenty organic results for best cars for resale value in the United States and read the bodies, not just the domains. Eighteen organic results came back, and not one of them is a government source.

Four of the eighteen trace back to a single Kelley Blue Book awards announcement: KBB’s own page at position 2, an industry outlet at 13 that reprints the press release word for word, a Nexstar wire item at 17, and a Forbes piece at 20 whose headline ends “Says KBB”. Two more carry an identical title — The Top 25 Cars That Hold Their Value Best — which is the headline of an iSeeCars study, one copy on Reddit and one on a West Virginia TV station’s website. So the page is not eighteen answers. Two studies are doing most of the work.

That would be fine — two good sources beat eighteen bad ones. What is not fine is what neither of them measures. Of the eighteen, eleven answered a plain fetch and ten served a readable article body; the other seven returned 403 and I make no claim about what is on them. Of the ten results that served a readable body, zero mention the IRS. Zero contain the phrase “cents per mile” or “standard mileage rate”. Zero mention the Truth in Mileage Act. And on Kelley Blue Book’s own awards page — the document the rest of the page copies — the words “mile”, “mileage” and “odometer” appear zero times.

That is a strange absence, because federal law says the odometer is the index of a car’s value, and the IRS publishes a number for exactly what one mile takes off it. In 2026 that number is 35 cents.

The two lists, and what they actually say

Both studies came out within a week of each other in March 2026, and both are honest work. Here is what each one found.

Kelley Blue Book, 2026 Best Resale Value Awards projects what a 2026 model will be worth after sixty months, as a percentage of its original MSRP. KBB’s own summary of the baseline: “the average new vehicle will be worth about 44.7% of its original sticker price after 60 months”, while the ten award winners “will return an average of 56.2% to their owners’ pockets” and “retain their value better than 95% of all other models.”

Rank	2026 model	5-year resale value (KBB)
1	Toyota Tacoma	63.0%
2	Toyota Tundra	59.9%
3	Toyota 4Runner	58.0%
4	Toyota GR Supra	56.0%
5	Mercedes-Benz G-Class	55.0%
6	Toyota Sienna	54.3%
7	Ford Maverick	54.1%
8	Chevrolet Corvette	54.0%
9	Porsche 911	53.9%
10	Ford Ranger	53.4%

iSeeCars did the opposite. It analyzed “over 950,000 5-year-old used cars sold from March 2025 to February 2026” and measured what actually happened, reporting an average five-year depreciation of 41.8%.

Rank	Model	5-year depreciation (iSeeCars)	Avg $ below MSRP
1	Porsche 718 Cayman	9.6%	$6,988
2	Porsche 911	11.1%	$15,533
3	Chevrolet Corvette	18.7%	$13,365
4	Toyota Tacoma	19.9%	$6,426
5	Toyota Tundra	21.2%	$8,746
6	Honda Civic	22.9%	$5,828

Now put the two side by side on the same truck. KBB says the Toyota Tacoma keeps 63.0% of its value, so it loses 37.0%. iSeeCars says the Tacoma loses 19.9%. Same model, same year of publication, both expressed against MSRP — and the answers are 17.1 percentage points apart. On the averages the gap is 55.3% versus 41.8%, which is 13.5 points, or about $4,725 on a $35,000 car.

Both numbers are correct. They answer different questions, and I will come back to why. First, the sentence that is missing from the page you actually read.

The assumption that is not on the page: 75,000 miles

KBB explains its method on the awards page itself. Residual values, it says, “are established by experienced automotive analysts who review the output from statistical models built upon millions of transactions”, using “a comprehensive data set that includes vehicle specifications, trims, sales volume, economic conditions, and auction results.”

Read that list again. Specifications, trims, sales volume, economic conditions, auction results. No mileage.

But KBB’s press release — the version republished verbatim by CBT News, which sits at position 13 on the same search results page — carries one extra sentence that the consumer-facing page does not:

Residual values reflect projected future auction values for vehicles in average condition with 75,000 miles at the end of a five-year lease or ownership period.

There it is. The award has a mileage assumption after all, and it is 75,000 miles over five years — 15,000 miles a year. I checked both of KBB’s award URLs: the string “75,000” appears zero times on each, and so does “mile” in any form. The assumption is real, it drives every percentage in the table, and it lives in the press kit rather than on the page.

Fifteen thousand a year is also the wrong number for most people, and the federal government measures how wrong. The Federal Highway Administration’s Table VM-1 for 2023 divides total vehicle-miles by registered vehicles and gets 11,026 miles per light-duty short-wheelbase vehicle — 11,106 across all light-duty vehicles, 11,408 across every motor vehicle on American roads.

	Miles per year	Miles over 5 years
KBB residual assumption	15,000	75,000
FHWA measured average (light duty, 2023)	11,026	55,130
Difference	+36.0%	+19,870 miles

So the ranking you are reading assumes you drive 36% more than the average American vehicle actually does. If you drive less than that — and most people do — every percentage in that table understates what your specific car will be worth. If you drive more, it overstates it. Either way the table cannot tell you, because it is a table about model years, not about you.

The confusion has a familiar source. “Twelve to fifteen thousand miles a year” is a figure for a driver, not a vehicle, and there are roughly 47 million more registered vehicles in the United States than there are licensed drivers to sit in them. I went through that arithmetic in detail in the piece on questions to ask when buying a used car, where the same mix-up makes every used odometer look low.

What Congress says a car’s value is indexed to

Here is the part that made me want to write this article. In 1972 Congress passed what became the Truth in Mileage Act, and the findings it wrote into the statute are still on the books today at 49 U.S.C. §32701:

Congress finds that — (1) buyers of motor vehicles rely heavily on the odometer reading as an index of the condition and value of a vehicle.

An index of the condition and value. That is the legislative reason the odometer is protected at all: not sentiment, not curiosity, but the fact that federal law treats the number on the dash as the proxy for what the car is worth. Everything downstream — the disclosure a seller must sign, the five-year retention requirement, the criminal penalties for rolling a dial back — exists because Congress decided that number carries the value.

That phrase appears zero times across the ten readable pages on the search results for best cars for resale value. The ranking industry and the statute are describing the same asset and have agreed to measure two different things about it.

What the IRS says a mile costs: 35 cents of 72.5

The other half of the answer is a document millions of Americans use every April without reading past the first line.

IRS Notice 2026-10 sets the 2026 standard mileage rate at 72.5 cents per mile for business use. Four pages later, in Section 4, it does something the headline never mentions:

For automobiles a taxpayer uses for business purposes, the portion of the business standard mileage rate treated as depreciation is 26 cents per mile for 2022, 28 cents per mile for 2023, 30 cents per mile for 2024, 33 cents per mile for 2025, and 35 cents per mile for 2026.

Thirty-five cents of the 72.5-cent rate is depreciation — 48.3% of it. Just under half of what the federal government says a mile costs you is not fuel, not tires, not insurance. It is the car quietly becoming worth less.

Tax year	Depreciation treated as taken, per mile
2022	26¢
2023	28¢
2024	30¢
2025	33¢
2026	35¢

That is a 34.6% increase in four years, which tracks a used-car market that never went back to where it was. The Bureau of Labor Statistics index for used cars and trucks is 29.4% above July 2019 as of July 2026, against 22.1% for new vehicles and 30.1% for all consumer prices — and it is still 13.6% below its July 2022 peak. Cars have not stopped getting expensive; they have stopped getting more expensive as fast.

And this is not a rule of thumb the IRS offers as guidance. Revenue Procedure 2019-46, section 4.04, makes it the tax basis of your vehicle:

Under § 1016(a)(2), a taxpayer must reduce the basis of an automobile used in business by the greater of the amount of depreciation the taxpayer claims for the automobile or the amount of depreciation allowable. If a taxpayer uses the business standard mileage rate to compute the expense of operating an automobile for any year, a per-mile amount (published by the IRS in an annual notice) is treated as the depreciation claimed by the taxpayer and the depreciation allowable for those years in which the taxpayer used the business standard mileage rate.

Treated as the depreciation claimed. If you deduct business miles at the standard rate, the IRS reduces your car’s basis by 35 cents for every one of them, whether or not you ever thought of it as depreciation. When you sell the car, that reduced basis is what your gain is measured against. The number is not advisory. It is already applied.

I keep the year’s rate and the mechanics of claiming it in a separate guide — IRS mileage rates 2026 — because the deduction is a different conversation from the one on this page. Here the rate matters for a reason that has nothing to do with taxes: it is the only official price tag anybody puts on a single mile.

The tax treatment described here is general information, not advice for your situation — talk to your accountant before acting on any of it.

The number that shows up twice

Now the two worlds can be put in the same sentence, and something odd happens.

KBB’s own example car is a $35,000 vehicle. At the average retention it publishes, 44.7% of MSRP after sixty months, that car loses $19,355 over five years.

Take the same five years measured in miles instead. The FHWA average is 11,026 miles a year, so 55,130 miles. At the IRS depreciation figure of 35 cents each, those miles cost $19,295.50.

Method	Five-year loss on a $35,000 car
KBB residual (44.7% of MSRP retained)	$19,355.00
IRS depreciation × FHWA average miles	$19,295.50
Difference	$59.50 — 0.31%

A residual-value analyst projecting auction prices off a sticker, and a tax rule built from cost surveys, land within a third of a percent of each other. They were never designed to agree. That they do is the best evidence I have found that the per-mile figure is not a bureaucratic fiction: it is roughly what a mile actually costs an American car owner, expressed in a way you can act on.

Which lets me put the entire resale-value industry into one line of arithmetic.

KBB’s headline benefit for choosing an award winner is $4,025 — the gap between 56.2% and 44.7% retention on a $35,000 car, which is where its own “more than $4,000” figure comes from. Divide $4,025 by 35 cents.

11,500 miles.

Choosing the single best value-holding vehicle in America, out of every model on sale, is worth about the same as not driving for one year. The lists are not wrong. They are just competing for a smaller prize than the odometer is.

Where the two studies disagree, and why it is not a contradiction

Back to the 13.5-point gap, because understanding it is what turns both studies into something usable.

KBB projects forward. It is pricing a 2026 model that will be sold in 2031, and its residuals are the numbers banks and leasing companies use to set lease payments today. A residual has to be conservative: if it is too high, the lender eats the difference when the car comes back. KBB is answering what should I assume this new car will be worth?

iSeeCars measures backward. It looked at cars registered in 2020 and 2021 and sold in 2025 and 2026, against an MSRP that it inflation-adjusted using BLS data, working from average asking prices. Those cars lived through the strangest used-vehicle market in living memory — the one that peaked in July 2022 with the CPI for used cars 50% above 2019. iSeeCars is answering what did the last five years actually do to these cars?

	KBB	iSeeCars
Direction	Projection, 2026 → 2031	Measurement, 2020/21 → 2025/26
Base	Nominal MSRP	MSRP adjusted to 2026 dollars (BLS)
Prices used	Projected auction values	Average asking prices
Sample	Residual Value Guide models	950,000 five-year-old cars
Mileage assumption	75,000 miles (not on the public page)	none stated
Average five-year loss	55.3%	41.8%

Neither is measuring the odometer. One assumes a mileage it does not print; the other does not mention mileage in its methodology at all. If you want the difference in one sentence: KBB tells you what to expect from a model, iSeeCars tells you what happened to a market, and neither tells you what will happen to your car.

Where they do agree is worth noting, because agreement across two hostile methods is a real signal. Both put Toyota trucks and Porsche sports cars at the top. iSeeCars finds trucks the strongest segment at 34.2% depreciation and hybrids close behind at 35.4%, with SUVs at 44.9% — and EVs worst at 57.2%, more than fifteen points off the overall average. KBB’s own list points the same way: its best-retaining electric car, the Tesla Model 3, sits at 35.0% retention, against 63.0% for the Tacoma.

If reliability rankings are also part of your decision, they disagree among themselves in exactly the same way, and for exactly the same kind of methodological reason — I took four of them apart in most reliable car brands.

So what does mileage actually do to a car’s value?

This is the question with no clean public answer, and I would rather say that plainly than invent a coefficient.

There is no federal dataset that prices a mile against resale value by model. What exists is the three things above, and they bound the problem from different sides:

• Federal law says the odometer is the index of condition and value (49 U.S.C. §32701).
• The IRS prices a business mile’s depreciation at 35 cents for 2026, and applies it to your basis whether you notice or not (Notice 2026-10, Rev. Proc. 2019-46 §4.04).
• The benchmark ranking is computed at 75,000 miles over five years, so its percentages are quotes for one specific level of use.

Put together, those give you something more useful than a ranking. They give you a way to convert your own driving into dollars. Thirty-five cents a mile is a national average built from cost data, not a valuation of your car — but as a working figure it beats guessing, and unlike a model ranking it responds to the thing you control.

Miles you drive per year	Depreciation at 35¢/mile	Over five years
5,000	$1,750	$8,750
8,000	$2,800	$14,000
11,026 (FHWA average)	$3,859	$19,296
15,000 (KBB assumption)	$5,250	$26,250
20,000	$7,000	$35,000

The spread between the top and bottom rows is $26,250 over five years. The spread between the best and average car on the KBB list is $4,025. That is the whole argument of this page in two numbers: how much you drive moves resale value roughly six times as much as which model you picked.

And the two are not independent. A high-mileage car of a strong-residual model still sells against low-mileage examples of the same model, which is precisely why the odometer reading and not the badge is what a buyer negotiates against — and why the seller has to certify it in writing.

The other half of the bill: what the car costs while you own it

Resale value is only one side of ownership cost, and the searches next to this one — cheapest cars to maintain, cars with lowest maintenance cost, most expensive cars to maintain — are asking about the other side. They deserve the same warning.

Maintenance-cost rankings are built from repair-shop and dealer data, which means they are compiled by the people who charge for the work. That is the same structural problem I measured on the maintenance searches: on the two big service-interval queries, 18 of 35 organic results were dealerships, repair chains, tire shops or parts sellers. The service schedule your car actually requires is in the manufacturer’s own maintenance booklet, which is free and which almost nobody opens — the argument is laid out in car maintenance schedule.

For this page the relevant point is narrower: maintenance cost, like depreciation, scales with miles rather than with years. An oil change is due at an interval measured in miles. Tires, brakes, belts and fluids all wear on distance. So the two halves of ownership cost move together, driven by the same variable, and the rankings for both are published as if that variable were fixed.

Fuel is the third piece, and it is the only one where a model choice really does dominate — the EPA publishes its own per-model numbers, which I went through in cars with the best gas mileage.

What to record if you want the answer for your own car

None of the above needs an app. It needs three numbers kept honestly over time, and most people do not have them when they need them, which is at the moment of sale, when a buyer or a dealer is quoting a figure and there is nothing to argue with.

The three are: the odometer at known dates, what you spent on the car, and the service that was actually done, with proof.

That last one is where the money is, and it is the part people reconstruct badly from memory. A car with a documented service history sells against a car without one, and the difference is not a rounding error. This is why the maintenance log matters at resale for exactly the same reason it matters at lease return — the charge is assessed against a standard, and standards are argued with evidence, a point I made at length in the piece on lease vs buy a car.

I built Magica around those three. It records trips automatically, so the odometer history builds itself instead of depending on you remembering to write a number down. It keeps a full maintenance log with scanned receipts attached, so the service history is a file and not a claim. And it produces reports on mileage, running costs and maintenance spend that export to PDF, which is the form a buyer, an accountant or a leasing company will accept.

All of it stays on the device. There is no account, no server of mine holding your driving history, and no sync through anything I control — Apple’s iCloud handles backup, and the cross-platform sync between iOS and Android is device-to-device. That is an architecture decision rather than a marketing promise: a log of everywhere a car has been is not data I want to be responsible for.

What it will not do is tell you what your car is worth. No app on your phone can, because that number comes from a market, not from your records. What the records do is make the conversation about that number an evidenced one.

The question the search engine asks that has no answer

One last thing worth flagging, because it appears directly under the search box.

The People Also Ask box on best cars for resale value includes the question “What is the $3,000 rule for cars?” I went looking for it. The string does not appear in Notice 2026-10, in Revenue Procedure 2019-46, in 49 U.S.C. §32701, or in 49 CFR part 580 — the federal texts that actually govern mileage, depreciation and odometer disclosure. There is no $3,000 rule in any of them.

It is a good illustration of the shape of this whole topic. The searches are full of thresholds and rules of thumb that sound official and are not, while the documents that are official — a 1972 congressional finding and an annual IRS notice — are sitting in plain sight, unread, saying something more useful than any of them.

FAQ

Which cars have the best resale value in 2026?

On Kelley Blue Book’s 2026 awards the top three are the Toyota Tacoma (63.0% of MSRP retained after five years), the Toyota Tundra (59.9%) and the Toyota 4Runner (58.0%). iSeeCars, measuring real five-year-old cars sold between March 2025 and February 2026, puts the Porsche 718 Cayman first at 9.6% depreciation, then the Porsche 911 at 11.1% and the Chevrolet Corvette at 18.7%. Both lists are built on the calendar, not on how far the car was driven.

How much does mileage affect car value?

No public dataset prices it per model, but two federal documents bound it. Congress writes at 49 U.S.C. §32701 that buyers rely on the odometer “as an index of the condition and value of a vehicle”, and the IRS sets the depreciation portion of the 2026 standard mileage rate at 35 cents per mile. At that figure, the difference between driving 5,000 and 20,000 miles a year is $26,250 over five years — roughly six times the $4,025 advantage KBB attributes to picking its best-retaining model.

Does mileage affect car value more than the model?

On those numbers, yes, and by a wide margin. The whole spread between the average car and the best-retaining car in America is about $4,025 on a $35,000 vehicle. Fifteen thousand extra miles a year, over the same five years, is worth about $26,250 at the IRS depreciation figure.

What mileage does Kelley Blue Book assume?

75,000 miles over five years, or 15,000 a year. The figure appears in KBB’s press release — republished by CBT News among others — and states that residual values reflect projected auction values “for vehicles in average condition with 75,000 miles at the end of a five-year lease or ownership period.” It does not appear on either of KBB’s public awards pages, where the words “mile” and “mileage” do not occur at all.

Why do KBB and iSeeCars give such different depreciation numbers?

Because they answer different questions. KBB projects forward what a 2026 model should be worth in 2031, against nominal MSRP, for lenders setting lease terms. iSeeCars looked backward at 950,000 five-year-old cars sold in 2025 and 2026, against inflation-adjusted MSRP, using average asking prices — in a used market still 29.4% above July 2019. The averages come out 55.3% versus 41.8%, a gap of 13.5 points.

What are the cheapest cars to maintain?

Maintenance-cost rankings are assembled from dealer and repair-chain data, so they come from the businesses that bill for the work — the same conflict that shows up on service-interval searches, where 18 of 35 organic results sell service or parts. The manufacturer’s own maintenance booklet is the free, authoritative source for what a specific car requires, and maintenance cost scales with miles driven rather than with years owned.

Do electric cars hold their value?

Not well, on both studies. iSeeCars puts EVs at 57.2% five-year depreciation, more than fifteen points worse than the 41.8% overall average and the weakest of every segment it measured. KBB’s best-retaining electric car, the Tesla Model 3, holds 35.0% of MSRP against 63.0% for the Toyota Tacoma. Hybrids are the opposite case: 35.4% depreciation, second only to trucks.

Does the IRS depreciation figure apply if I do not deduct mileage?

No. The 35-cent figure is a tax rule, and Revenue Procedure 2019-46 applies it to basis only for taxpayers who use the business standard mileage rate. Its usefulness here is different: it is the one official, annually updated price for what a mile of driving costs, and it is built from real cost data rather than from opinion. Check your own tax position with your accountant.

What I would actually do with this

Pick the model on reliability and on what the car has to do, use the resale rankings as a tiebreaker rather than a decision, and then pay attention to the variable that turns out to be worth six times as much: how far you drive it, and whether you can prove what was done to it.

The rankings are a forecast about a model. The odometer is a fact about your car. Congress said so in 1972, and the IRS puts a price on it every December.