---
title: "The Best Time to Buy a Used Car, According to 259 Months of Price Data"
description: "Federal price-index data covering two decades shows used-car prices peak in August and bottom in December — and that the reliable fall happens in September. What timing is really worth."
url: "https://baronauto1.com/buying-guides/best-time-to-buy-a-used-car/"
type: "article"
published: "2026-08-27"
modified: "2026-08-27"
site: "Baron Auto"
disclaimer: "This site is under new ownership and is not affiliated with Baron Auto Emporium dealership."
---

# The Best Time to Buy a Used Car, According to 259 Months of Price Data

> Federal price-index data covering two decades shows used-car prices peak in August and bottom in December — and that the reliable fall happens in September. What timing is really worth.

*Used Car Buying · 14 min read · 3,221 words*

## The short version

- There is a real seasonal pattern in used-car prices, and it is measurable. It is also smaller than almost everyone selling you advice implies.
- Across 259 months of the federal used-vehicle price index, prices climb from February to a peak in August, then fall through the autumn and bottom out in December.
- September is the one genuinely reliable month. The index fell in 16 of the 18 Septembers in this sample — the most consistent seasonal move in the series.
- December is the cheapest month, but by then the fall has already happened. December’s own move is only −0.22%, and it fell in just 11 of 18 years.
- The whole distance between the dearest month and the cheapest is 3.63%. On most cars that is less than the gap between two sellers on the same afternoon.
- Borrowing costs move far more than the calendar does. The 48-month rate went from 4.17% to 7.47% in a decade, which swamps a 3.63% seasonal swing.

Every guide to this question gives you the same three answers: buy at the end of the month, buy at the end of the quarter, buy in December. They are repeated so often that they have stopped sounding like claims and started sounding like weather.

None of them is sourced. So rather than repeat them, this article asks a narrower question that can actually be answered: across two decades of federal price data, which months are used cars cheapest in, and by how much?

The answer is more useful than the folklore, and less dramatic.

![An overhead view of a large tarmac lot marked out in numbered parking bays, perhaps a third of them filled with cars in dark and light colours](https://baronauto1.com/assets/photos/auction-lane-1280.webp)

*Supply is the thing that actually moves used-car prices, and it arrives on a calendar. The autumn is when the lots fill up, and it is the only part of the year where the price effect is large enough and repeatable enough to plan around.*

## What the measurement actually is

The figures in this article come from one series: the Consumer Price Index for used cars and trucks, published monthly by the Bureau of Labor Statistics. It is a price index, which means it tracks what a constant basket of used vehicles costs from one month to the next.

That distinction matters more than it sounds. The index can tell you whether this month is dearer than last month. It cannot tell you what any particular car should cost, and nothing here is a valuation. If you want to know what the car in front of you is worth, the index is the wrong instrument entirely.

Two decisions were made in handling it, and both are worth stating plainly because they change the answer.

**The series used is the one that has not been seasonally adjusted.** BLS publishes two versions of most indexes: a raw one and a seasonally adjusted one. The adjusted version exists precisely to strip out repeating calendar effects so that economists can see the underlying trend. Using it here would have been circular — it is the seasonal signal that has been removed. This article uses the raw series, which is the only one that can answer the question.

**The 2020 to 2022 period is reported separately.** Semiconductor shortages halted new-car production, buyers moved into the used market, and used-vehicle prices rose and then fell in a way that had nothing to do with what month it was. Leaving those three years in would let a chip shortage masquerade as a season. So every figure below was computed twice: once across the whole span, and once with 2020, 2021 and 2022 removed.

The reassuring result is that it barely matters. Both versions name the same dearest month and the same cheapest month, and the distance between them is 3.66% across the full span against 3.63% with the shock years removed. The seasonal pattern is robust to the largest disruption the used-car market has had this century, which is the strongest evidence available that it is real.

## The shape of a year

Rather than twelve separate percentages, it is easier to read the year as a single curve: take the typical change for each month and chain them together into one notional year.

**Figure: The shape of a year in used-car prices**

A line running across the twelve months of a notional year, built by chaining the median month-over-month change in the federal used-vehicle price index. It climbs from February to a peak in August, falls steeply through the autumn, and bottoms in December.

Built from 259 months of the Bureau of Labor Statistics used-vehicle price index, with the 2020 to 2022 price shock excluded. The gap between the dearest month and the cheapest is 3.63 per cent — real money on a used car, but far less than the spread between two dealers on the same afternoon. The index tracks what a constant basket of used vehicles costs, so it says whether this month is dearer than last month, never what any particular car should cost.

The pattern is a long climb and a short drop. Prices drift upward from February, gathering pace through the spring, and keep rising until August. Then the floor gives way. September, October and November are all sharply negative, and by December the index has given back everything it gained and a little more.

The obvious question is why. The honest answer is that the index does not explain itself — it records what happened, not why. But the shape is consistent with the supply story that people in the trade describe: spring is when tax refunds arrive and buyers turn up, and autumn is when new-model-year introductions push a wave of trade-ins and ex-fleet cars onto the market. Demand leads in the spring; supply catches up in the autumn.

Treat that as the plausible mechanism rather than a proven one. The measurement stands on its own regardless.

## September is the month that actually repeats

A median tells you the usual size of a move. It does not tell you how often that move turns up at all, and for a decision you are going to make once, the second question matters more.

**Figure: How often the used-vehicle price index fell, by month**

Twelve bars counting, for each calendar month, the number of years in which the federal used-vehicle price index fell rather than rose.

September is the month to watch: the index fell in 16 of the 18 Septembers in this sample, which is the most consistent seasonal move in the series. March is its mirror, falling in only 3 of 19. A median tells you the usual size of a move; this tells you how often it turns up at all, and the two answer different questions.

September is the standout. The index fell in 16 of the 18 Septembers in the sample — not a tendency but close to a rule. October follows at 15 of 18 and November at 14 of 18. Nothing in the first half of the year comes close: March fell in 3 years out of 19, June in 3 out of 19.

So the single most defensible statement this data supports is not about December at all. It is that *the used-car market turns in September*, and it turns almost every year.

If you have a choice about when to start shopping, starting in early September puts you at the top of the slide rather than at the bottom of it, with the widest selection still on the lots and three months of softening ahead of you.

## Why “buy in December” is half right

December is genuinely the cheapest month. The folklore is not wrong about the destination.

It is wrong about the mechanism, and that changes what you should do. The usual explanation is that December is cheap because salespeople are chasing year-end targets. If that were the whole story, December itself would show a large drop. It does not: December’s own median move is −0.22%, the smallest change of any month in the falling half of the year, and it fell in only 11 of 18 years — barely better than a coin toss.

December is cheap because September, October and November were cheap first. It is the bottom of a slope, not a cliff edge. Someone who waits from August to December captures the whole decline; someone who decides in late November to hold out for December is haggling over the last 0.22% while the selection thins out around them.

The practical version: the autumn is the opportunity, and December is simply the last month of it.

## The months to avoid are clearer than the months to target

The strongest signal in this data is arguably not when to buy but when not to.

March is the worst month to be shopping. Its median move is +0.79%, and it rose in 16 of the 19 Marches observed. April, May, June and July are all positive too. A buyer who shops across the spring is walking up an escalator: not steeply, but consistently, and in the same direction every year.

The cumulative effect is the part people miss. No single spring month looks alarming on its own — half a percent here, six-tenths there. Chained together from February to August they account for the entire 3.63% climb. The spring does not have one bad month; it has six mediocre ones in a row.

If your timing is flexible at all, avoiding March through July is a more reliable decision than targeting any particular week.

## What the calendar is actually worth

Here is where this article parts company with most of what is written on the subject.

The entire distance from the dearest month to the cheapest is 3.63%. That is the whole prize. It is real money and it is worth having, but it needs to be held against the other levers available on the same purchase.

The spread between two sellers listing comparable cars in the same week is routinely wider than 3.63%. So is the difference between the advertised price and the out-the-door price once fees are added, which is why our guide to [asking for the out-the-door price](https://baronauto1.com/buying-guides/out-the-door-price/) is a more valuable habit than watching the calendar. So, often, is what a pre-purchase inspection turns up.

None of that makes the seasonal effect fake. It makes it a tiebreaker rather than a strategy. If you are choosing between buying in April and buying in October and everything else is equal, October. If you are choosing between a well-documented car in April and an unknown one in October, that is not a timing question and the calendar should not decide it.

## The lever that dwarfs the calendar

If you are borrowing, the interest rate moves far more money than the month does.

**Figure: Advertised auto loan rates at commercial banks, a decade apart**

Paired bars comparing the average finance rate on 48-month and 60-month car loans at commercial banks ten years ago against the most recent published figure.

Federal Reserve G.19 figures, which are the rates banks advertise on new-car loans. A used car costs more to finance than this and a private-party loan more again, so read these two bars as the floor beneath your quote rather than as the quote. The floor itself moved from 4.17 to 7.47 per cent on the 48-month series, which is most of why a monthly payment that felt normal a decade ago buys a cheaper car today.

The Federal Reserve publishes the average finance rate on car loans at commercial banks. On the 48-month series it went from 4.17% to 7.47% over the decade to the most recent reading. The 60-month series moved from 4.11% to 7.14% over the same period.

Those are advertised new-car rates, which matters: a used car is financed at more than this, and a private-party loan at more again. Read them as the floor beneath your quote rather than as your quote. But the direction is unambiguous, and the size of the move is several times the entire seasonal spread.

The average amount financed moved too, from $28,140.06 to $42,503.54 across roughly the same decade, which is the highest reading in that series. A larger balance at a higher rate is why a monthly payment that felt ordinary ten years ago now buys a noticeably cheaper car.

The conclusion follows directly: shopping the loan is worth more than shopping the month. A borrower who improves their rate has done more for their total cost than a borrower who waited from August to December. Doing both is better than doing either, but if you only have the attention for one, it should be the financing.

## What a season is worth against what a market is worth

The most useful thing this dataset does is calibrate the reader’s sense of scale, and it does that by containing one enormous counter-example.

Between June 2020 and June 2021 the used-vehicle index went from 136.915 to 198.855 — a rise of 45.24% in twelve months, the largest year-on-year move anywhere in the series. It peaked at 213.683 in July 2022. For comparison, the lowest reading in the whole span is 121.061, in March 2009, at the bottom of the financial crisis.

Set the seasonal effect beside that. The entire annual swing is 3.63%. The 2021 disruption was more than twelve times larger and lasted years rather than months.

Two things follow. The first is that *which market you are buying in matters far more than which month*. A buyer with perfect seasonal timing in 2021 paid vastly more than a buyer with terrible timing in 2019, and no amount of waiting for December would have closed that gap. The second is a caution about the advice industry: a great deal of used-car guidance written between 2021 and 2023 was describing a semiconductor shortage while believing it was describing the used-car market.

This is also the reason the seasonal figures in this article exclude those years rather than averaging through them. A 45.24% shock sitting inside a dataset will dominate any month-of-year average computed across it, and the resulting “seasonal pattern” would mostly be a record of when the chip shortage happened to bite.

## The tactics the data cannot see

End of month, end of quarter, the last hour of a bank holiday weekend — none of these show up in a monthly index, because a monthly index has no resolution below a month. This article cannot confirm them and cannot refute them.

What can be said is narrower and still useful. Those tactics are all claims about an individual seller’s incentives on a particular day. They are not claims about the market. That means their value varies enormously between one seller and the next, they cannot be verified in advance, and they are unavailable to you when buying privately, which is a large share of the market.

The seasonal pattern is the opposite kind of fact. It is weaker per transaction and far more dependable, because it describes the whole market rather than one desk. Use the tactics if you enjoy them. Do not plan around them.

## Model-year timing, and the trade-in you bring with you

Autumn is also when the new model year lands, and that cuts both ways.

On the buying side it is the mechanism behind the price fall — outgoing models get discounted, lease returns and fleet cars arrive, and the supply of two-to-four-year-old vehicles thickens. On the selling side it is the worst possible moment to be handing over a trade-in, because your car is being valued into the same glut.

If you are trading in, the timing question has two answers pulling against each other and you cannot optimise both. Our guide to [how a car trade-in works](https://baronauto1.com/trade-your-car/how-does-a-car-trade-in-work/) covers the mechanics of keeping the two transactions separate, which is the usual advice and is especially worth following in the autumn, when the number you gain on the purchase can quietly be taken back on the trade.

## The practical sequence

Putting the above in the order it should actually be done:

- **Start looking in early September if you can.** Not because September is cheapest, but because it is the reliable turn and the selection is still wide.
- **Sort the financing before the car.** Get an approval in hand from a bank or credit union so the rate is a fixed input rather than something negotiated at a desk by someone earning on it.
- **Do not stretch the timeline to reach December.** The last month of the slide is worth 0.22%. It is not worth three months of a car you cannot rely on.
- **Avoid committing in March through July if the choice exists.** This is the more reliable half of the finding.
- **Let the individual car override all of it.** A 3.63% seasonal spread does not survive contact with one deferred timing-belt service.

## The month is the smallest variable in the purchase

Everything above concerns the market. None of it concerns the car, and the car is where the money actually is.

A vehicle with an open safety recall, an inconsistent odometer record, or a title branded in another state is a worse purchase in December than a sound car is in March, by a margin that makes the seasonal question irrelevant. Those are all checkable before you go, from the VIN alone, and our guide to [checking a used car’s history](https://baronauto1.com/vehicle-history/how-to-check-a-used-cars-history/) sets out the free federal sources in the order worth using them.

If you want the decode, the title-brand record and the open recalls in one pass, you can [run the VIN through CarCheckerVIN](https://carcheckervin.com) before you spend an afternoon on a car that the record would have ruled out. Do that first. Then worry about the month.

## Common questions

### What is the single best month to buy a used car?

December, on this data — it is the low point of the year in the federal used-vehicle price index. But the margin over November is small, and December’s own median move is only −0.22%. September through December is the better way to think about it than any one month.

### How much money does buying at the right time actually save?

The gap between the dearest month and the cheapest is 3.63% in the index. That is the ceiling on what pure timing can be worth, and it assumes you would otherwise have bought at the annual peak. It is usually smaller than the difference between two sellers.

### Is it true that you should buy at the end of the month?

This data cannot say. A monthly index has no resolution finer than a month, so the claim is neither confirmed nor refuted here. It is a claim about one seller’s incentives rather than about the market, and it does not apply at all to private sales.

### Why is September the month that keeps coming up?

Because it is the most consistent. The index fell in 16 of the 18 Septembers in this sample, a higher hit rate than any other month. It is also the largest single monthly fall, at a median of −1.41%.

### Should I wait until December if I need a car in April?

Almost certainly not. Eight months of arrangements, hire costs or an unreliable current car will cost more than 3.63% of a used-car purchase. Timing is a tiebreaker for buyers who are already flexible, not a reason to become flexible.

### Does the 2021 price spike make this data unreliable?

It was checked. Every figure was computed both across the whole span and with 2020, 2021 and 2022 excluded. Both versions give the same cheapest month, the same dearest month, and a spread of 3.66% against 3.63%. The pattern survives the shock.

### Does this apply to the car I am looking at?

Only loosely. The index describes a national basket of used vehicles, not a model, a trim, or a region. A specific car’s condition, history and mileage move its value far more than the month does, which is why the vehicle record is worth more attention than the calendar.

## Sources and further reading

- [BLS Consumer Price Index (used cars and trucks)](https://www.bls.gov/cpi/)
- [Federal Reserve: finance rate on 48-month new car loans](https://fred.stlouisfed.org/series/TERMCBAUTO48NS)
- [Federal Reserve G.19 consumer credit release](https://www.federalreserve.gov/releases/g19/current/)
- [FTC used car buying guide](https://consumer.ftc.gov/articles/buying-used-car-dealer)
- [FTC Used Car Rule](https://www.ftc.gov/business-guidance/resources/dealers-guide-used-car-rule)
- [NHTSA recall lookup](https://www.nhtsa.gov/recalls)
- [NHTSA Office of Defects Investigation complaint database](https://www.nhtsa.gov/nhtsa-datasets-and-apis)
- [NMVTIS (US Department of Justice)](https://vehiclehistory.bja.ojp.gov/)
- [CFPB auto loan resources](https://www.consumerfinance.gov/consumer-tools/auto-loans/)

Recall, complaint and safety-rating figures on this page were retrieved from the federal databases above on August 19, 2026. Federal data changes — re-check any VIN before you rely on it.

---

*This site is under new ownership and is not affiliated with Baron Auto Emporium dealership.*

Canonical source: https://baronauto1.com/buying-guides/best-time-to-buy-a-used-car/