Polymarket Slippage 2026: 897 Books Measured
Polymarket slippage measured on 15 October 2026 across 897 active order books: the quote you click is good for a median of $197 before the price moves.
Affiliate disclosure: contains affiliate links; we may earn a commission at no extra cost to you. 18+ only — see BeGambleAware.org. Last updated 15 October 2026.
Key takeaways
- Polymarket slippage is not in the spread — it is in the depth behind it. Across 897 actively traded books the median amount available at the best ask was $197, and in 41.9% of them it was under $100. Everything above that walks up the book.
- A $1,000 market order fills at the quoted price in only 32.8% of those markets. The median cost of the rest was 2.29% of the stake; at the 90th percentile it was 134%.
- Fees are the bigger cost at $1,000 and slippage is the bigger cost at $10,000. The crossover sits between the two: slippage beat the fee in 33.5% of markets at $1,000 and 65.2% at $10,000.
- Cheap outcomes have the dearest execution. Under 2¢, a $1,000 buy moved the average fill 296% above the quote — and the taker fee is also at its highest share of stake there, not its lowest.
- Nearly half the exchange cannot take a $1,000 order at all. Of 9,286 two-sided books, 4,074 could not absorb $1,000, and 3,436 held under $100 of asks in total.
On this page
Every guide to Polymarket slippage tells you the same three things: use limit orders, split large orders, check the book first. None of them tells you the number that decides whether any of that matters — how much you can actually buy before the price moves.
That gap is expensive. The only figure anywhere in the search results is an asserted "5 to 10 cents on thin markets" in a bot tutorial. Get it wrong on a 6¢ contract and the fill costs eleven times the quote.
So we read the whole book. Every figure below comes from 12,667 live order books pulled from Polymarket's public CLOB and Gamma APIs in the eight minutes to 19:02 UTC on 15 October 2026, walked one price level at a time against the exchange's own published fee formula. We checked the documentation afterwards, and it contradicts itself in one place we flag below.
What a Polymarket market order actually costs
Of the 12,667 books, 9,286 had both a bid and an ask and could be priced at all. Narrowing to the markets a trader actually meets — at least $1,000 of 24-hour volume, priced between 5¢ and 95¢ — leaves 897 books. Those are the population for every figure in this section.
For each one we simulated a buy: take the cheapest ask, exhaust it, move up a level, repeat until the order is spent — which is exactly what Polymarket's own developer docs tell you to do to derive a maximum price.
Slippage is the average fill price against the best ask, as a percentage of the stake. 897 actively traded books priced 5¢-95¢, read 15 October 2026.
Read the first column before the others. At $100 you are usually fine and at $1,000 you usually are not — two markets in three move against you on a four-figure order. The medians stay small because the deep half of the sample absorbs anything, which is precisely why a median is the wrong number to plan with. The 90th percentile is the one that empties an account.
The $100,000 row is the one that reframes the exchange. 47% of these actively traded, mid-priced books cannot absorb a single $100,000 order at all, and among those that can, the median fill lands at more than double the quoted price. Institutional size is available in a few dozen books, not across the platform.
The spread looks tight because the spread is not the problem
Here is why this cost stays invisible. The median spread across those 897 books was exactly 1.00¢, and 62.7% of them quoted a spread of one cent or less. By the usual screening metric, Polymarket looks like a tight market.
It is tight. It is just not deep. The spread describes the price of the first share; the depth behind it describes the price of the thousandth.
A one-cent spread on a 50¢ contract is a 2% round-trip cost on the first $197. On the next $800 it is whatever the next price levels charge, and nothing on the screen shows you that until the confirmation does.
The wider exchange is thinner still. Across all 9,286 two-sided books, 4,074 could not fill a $1,000 order at any price and 3,436 held less than $100 of asks in total. Those are live, open, tradeable markets with order books that function as decoration.
The $197 quote: how much you can buy before the price moves
If you take one number from this article, take that one. The median book held $197 at its best ask — your no-slippage budget in a typical actively traded market.
The distribution is what makes it usable, because it is badly skewed. In 41.9% of these markets you had under $100 at the quote; in 9.0% you had over $10,000. Liquidity is concentrated, and 24-hour volume predicts where it sits almost perfectly.
The jump from the first row to the second is the whole trading strategy. Moving from a market doing $5,000 a day to one doing $50,000 a day cut the median cost of a $1,000 order from 5.43% to 0.40% — a thirteen-fold saving available for free, before any order-type cleverness.
At the top end the depth is genuinely institutional. The House-control market after the 2026 midterms held $613,214 at a single price level, and the Champions League outright on Paris Saint-Germain held $35,693. In books like those, slippage is not a consideration at retail size at all.
Slippage or fees: which one is actually taking your money
This is the comparison nobody makes, and it has a clean answer because both costs can be expressed in the same unit: cents per share.
Polymarket's taker fee is fee = C × feeRate × p × (1 − p), where C is the number of shares and p the price. Per share, that is feeRate × p × (1 − p) — so a crypto market at 50¢ charges 0.07 × 0.5 × 0.5 = 1.75¢ a share. Slippage is simply your average fill minus the best ask, in the same cents. We computed both for every market where the live fee schedule was readable.
The crossover sits between $1,000 and $10,000. Below it, the fee is the larger cost in two markets out of three and fee-shopping by category is rational. Above it, slippage is the larger cost in two out of three, and at the 90th percentile it is 34 times the fee — at which point the category of the market matters far less than the depth of the book.
That has a direct consequence for how you read our Polymarket fees guide. The fee schedule is worth optimising for small, frequent positions. For size, the fee is a rounding error against execution, and the right question is which book you are standing in.
The longshot trap: cheapest price, dearest execution
A 2¢ contract looks like the cheapest thing on the exchange. It has the most expensive execution on the exchange, and it is punished twice over.
The first penalty is arithmetic. Prices move in ticks, and a tick is a fixed number of cents against a variable price — so the same one-cent step that costs 2% of a 50¢ stake costs 20% of a 5¢ stake. Polymarket does mitigate this: sub-5¢ markets quote on a 0.1¢ tick rather than a penny, which is why the 2¢-5¢ band shows a 3.2% tick cost while the 5¢-15¢ band shows 7.7%. The 5¢-15¢ band is where the penny tick bites hardest, and it gets no relief.
The second penalty is the fee, and it runs the opposite way to intuition. As a share of what you stake, the taker fee is feeRate × (1 − p) — largest on the cheapest contracts. A 2¢ crypto longshot pays 3.98% of stake in fees where a 90¢ favourite pays 0.42%. The fee formula is symmetric in dollars and deeply asymmetric in what it costs you.
Put together: at the cheap end you pay the most in fees, cross the widest relative tick, and meet the thinnest books. A tweet-count market at a 6¢ ask with $10,944 of daily volume cost 1,103% above the quote on a $10,000 buy — and so did real sports: a Dota 2 game winner at 6¢ with $48,876 of daily volume came in at 712%.
Which categories fill, and which do not
Category turned out to matter more for execution than it does for fees, and in the opposite direction to the fee table.
Fee rates are the ones the markets themselves returned, and they vary inside a category — see below. Read the rate from the market, not from the category.
Sports is the best-executing category on the exchange — more than half of its actively traded books filled a $1,000 order without moving the price — and 29% of its markets also sit on 0.03, the cheapest non-zero fee rate live anywhere on the exchange. Crypto is the inverse: the highest rate at 0.07, and only 12.7% of books taking $1,000 at the quote.
Weather is the outlier worth naming, because not one of its 65 actively traded books absorbed $1,000 at the quoted price and the median fill landed 210% above it. Daily temperature markets settle fast and nobody makes markets in them.
The round trip is the number that matters
Everything above prices getting in. Most positions have to come out, and the cost of a round trip is not twice the slippage — it is the whole distance between the ask you buy at and the bid you sell into, across both books.
We simulated it: buy $1,000 of the YES side, then immediately sell every share back into the live bids.
Across 814 actively traded books, the median round-trip cost was 8.04% of the stake. The 25th percentile was 3.33%, the 75th 30.11%, and the 90th 67.82%. Buying and instantly reselling a $1,000 position in the typical active market destroys $80 before the event has resolved anything.
That figure reframes what an "edge" has to be. If your model says a contract is 3 points mispriced and the round trip costs 8 points, the trade loses money even when you are right — unless you intend to hold to resolution, which is the only exit that avoids the second leg entirely.
It also explains why Polymarket pays people to post the quotes. The maker-rewards programme exists because this depth has to be bought, and the 20.8x spread in what the same capital earns across markets maps onto the same thinness measured here.
What Polymarket's own docs do and do not tell you
Two things stand out from reading the documentation against the live API on the same afternoon.
First, Polymarket's docs never use the words "slippage" or "price impact" anywhere in the order-placement pages. They do describe the mechanism exactly — the developer guide tells you to walk the asks upward, accumulating price × size until your target amount is reached, and use the last price as your maximum. The method is published and the measurement is not, which is roughly how this article came to exist.
The order types are the lever the docs do give you, and they are worth knowing precisely.
A market order carries a maxPrice that "prevents a BUY from crossing a higher price", which is the slippage cap the interface does not call a slippage cap. Minimum order size was 5 shares in every market we sampled, and ticks were 1¢ or 0.1¢.
Second, and this is a genuine discrepancy: the published fee table does not match the live fee schedules. The docs list four taker rates — 0.07, 0.05, 0.04 and zero. The markets return five. Among the 1,000 highest-volume markets we read schedules for, 143 sports markets charge 0.03, not the documented 0.05, under two schedule versions the docs do not mention at all (one of them named for NFL and college football in October 2026). That is 29% of sports markets and 20.7% of sports 24-hour volume paying 40% below the published rate. The rebates differ too: 0.15 on one, 0.25 on the other.
Nothing here is a problem for a trader — it is a discount — but it means the fee you will pay is a property of the individual market, not of its category, and the only authoritative source is the market's own feeSchedule field. We confirmed the 89 geopolitics markets with no schedule at all are genuinely fee-free, exactly as the docs say.
Six ways to pay less for a fill
The standard advice is not wrong, it is just unquantified. Here is the same advice with the thresholds this data supports.
1. Treat $197 as the default no-slippage budget. In an unfamiliar market, assume the quote covers about $200 and check the book before exceeding it. In 41.9% of markets it covers less than $100.
2. Use 24-hour volume as your liquidity screen. It is the single best predictor available on the page. Above $10,000 of daily volume, the median $1,000 order costs 0.40%; below it, 5.43%.
3. Post, do not take. Makers are never charged fees on Polymarket — every live schedule we read carried takerOnly — so a resting GTC order saves the fee and the slippage. The price of that is uncertainty about getting filled.
4. Size the first leg to the depth at the quote, not to your conviction. FAK does this mechanically: it takes what is available now and cancels the rest, which converts a bad fill into a smaller good one.
5. Be suspicious of anything under 15¢. The 5¢-15¢ band carries a 7.7% tick cost, a 3.6% fee share and a 35% median slippage on $1,000. If you want a longshot, resting a limit order is close to the only sane route in.
6. Price the exit before the entry. An 8% median round trip means a position you may need to close early must clear roughly four times the hurdle of one you will hold to resolution.
Check the book before you size the trade
Every figure here is a 15 October 2026 snapshot, and depth changes by the minute. Read the live order book on the market itself.
Visit Polymarket.com →Affiliate link. 18+ only. Polymarket restricts access from the United States, Canada and other listed jurisdictions. Prediction markets carry a substantial risk of loss. Nothing here is financial advice.
My verdict
Polymarket's headline liquidity is real and it is concentrated in a few dozen books. Outside those, the exchange is a tight-spread, no-depth market where the quoted price is good for a couple of hundred dollars and the second price level is the one you actually trade at.
The practical consequence is that execution, not fees, is the cost to manage above about $1,000 — and it is the one cost no page on the platform or in the search results will quantify for you. A trader who screens on 24-hour volume, posts rather than takes, and refuses to market-order into anything under 15¢ has removed most of it without any special tooling.
What would change my mind is the depth at the top. A book holding $613,214 at one price level is a different exchange from the one in our median row, and if that profile spreads down the volume curve these numbers date quickly. They are one afternoon's snapshot — worth re-measuring before you trade on them.
Related guides
- Polymarket Fees 2026: What You Actually Pay — the fee side of the same trade, category by category, including the US app's separate schedule.
- Polymarket Maker Rewards 2026: $852 a Market — what the exchange pays to have the depth this article measures, and where it pays most.
- How Does Polymarket Work? Explained for 2026 — the YES/NO share mechanics behind every order book here.
- Polymarket Perps 2026: 88 Markets, 50x — the leveraged side, with its own funding, margin and liquidation maths.
- Polymarket Deposit Fees 2026: The Real Cost — what it costs to get collateral in before any of this applies.
- Polymarket vs Kalshi 2026 — the regulated alternative, and how its market structure differs.
Polymarket slippage FAQ
Does Polymarket have slippage?
Yes, although Polymarket's own documentation never uses the word. Its markets are a central limit order book, so a market order fills against whatever is resting and walks up the price levels until it is complete. Measured across 897 actively traded books on 15 October 2026, a $1,000 buy filled entirely at the quoted price in only 32.8% of them.
How much can you bet on Polymarket without moving the price?
A median of $197 in an actively traded market — that is how much was available at the best ask across the books we measured. In 41.9% of them it was under $100, and in 9.0% it was over $10,000. The largest book we found held $613,214 at a single price level, so the answer varies by three orders of magnitude and has to be read from the specific market.
Is slippage or the fee the bigger cost on Polymarket?
It depends on size, and the crossover is between $1,000 and $10,000. At $1,000 the median slippage was 0.215¢ a share against a median fee of 0.770¢, and slippage was the larger cost in 33.5% of markets. At $10,000 median slippage rose to 1.698¢ a share — 2.33 times the fee — and it was the larger cost in 65.2% of markets.
Why is slippage so bad on cheap Polymarket contracts?
Three reasons compound. Prices move in fixed ticks, so one tick is a far larger share of a 5¢ stake than a 50¢ one. The taker fee as a share of stake is feeRate × (1 − p), which is highest at low prices. And the books are thinnest there. Under 2¢, the median $1,000 buy filled 296% above the quote; between 5¢ and 15¢ it was 35%.
How do you avoid slippage on Polymarket?
Post a limit order rather than taking one. A resting GTC order pays no fee at all — every live fee schedule we read is taker-only — and no slippage, at the cost of not knowing whether you will be filled. If you must take, FAK takes only what is available immediately and cancels the remainder, and a market order's maxPrice caps how far it may cross.
Which Polymarket markets are the most liquid?
Sports, by a wide margin on execution: 53.1% of its actively traded books absorbed a $1,000 order at the quoted price, against 24.9% for politics and 12.7% for crypto. Weather was the worst of the categories we measured, with none of its 65 active books taking $1,000 at the quote. Twenty-four-hour volume predicts depth better than category does.
What does it cost to buy and sell on Polymarket straight away?
A median of 8.04% of the stake. We bought $1,000 of the YES side in 814 actively traded books and sold every share straight back into the live bids: the 25th percentile lost 3.33%, the 75th 30.11% and the 90th 67.82%. That round-trip cost is the real hurdle any short-term edge has to clear.
Sources
Order books for 12,667 live markets — every order-book market inside Polymarket's 400 highest-volume active events — were read from the public CLOB and Gamma APIs in the eight minutes to 19:02 UTC on 15 October 2026. 9,286 had two-sided books and could be priced; 897 of those had at least $1,000 of 24-hour volume and a price between 5¢ and 95¢, and are the population for the headline figures. Slippage is our own level-by-level walk of each ask side, buying the first listed outcome token; fees are the published formula applied to each market's own live fee schedule. Depth moves by the minute and these figures are one snapshot. Nothing here is financial advice.
- Polymarket docs: fees — the fee = C × feeRate × p × (1 − p) formula, the four published category rates and "makers are never charged fees".
- Polymarket docs: create an order — GTC, GTD, FOK and FAK definitions, maxPrice, tick-size rules and the ask-walking method for estimating a marketable price.
- CLOB API: /book — the full bid and ask ladders, tick size and minimum order size this article walks.
- Gamma API: /markets — 24-hour volume, best bid and ask, and the per-market feeSchedule that disagrees with the docs.
- sailgp.com: slippage in prediction markets — the best-ranking general explainer, and a page with no measured figure in it.
- dev.to: Polymarket bot order-book monitoring — the source of the "5 to 10 cents on thin markets" estimate this run replaces with a distribution.