Wash trading is when one person, or a group working together, buys and sells the same NFT between wallets they control. Nothing really changes hands, but the sale still shows up as genuine volume on the marketplace. The point is usually to make a quiet collection look busy, to lift a floor price before selling to a real buyer, or to farm a marketplace’s token rewards. Most of it can be spotted without special software. Look for the same two wallet addresses trading an item back and forth, sale prices far above everything else in the collection, a volume spike that arrives with no new owners, and buyer and seller wallets that were funded from the same place. Research published in the journal Research Policy in 2026 found patterns consistent with wash trading in about 38 percent of NFT trades and 60 percent of traded value across three major marketplaces.
If you have ever opened a collection page, seen a big green volume number, and bought in on the strength of it, this is the guide that explains why that number is often not what it looks like.
What wash trading actually is
A wash trade is a sale with no change in beneficial ownership. Wallet A sells a token to Wallet B for 5 ETH, and the same person holds the keys to both. The 5 ETH goes from one pocket to the other. The NFT moves the other way. At the end of it, the trader owns exactly what they owned before, minus gas and marketplace fees.
The reason it works as a manipulation is that a blockchain records addresses, not people. Ten wallets controlled by one person look like ten separate traders to anyone reading the chain, and to the marketplace’s own charts. There is no identity layer that says these are the same hands.
The practice is old and it is banned in regulated markets. What is different about NFTs is that the barrier to doing it is close to zero. Anyone can create an unlimited number of wallets in a few minutes, at no cost, with no verification.
Why people do it
Farming token rewards
Some marketplaces have paid traders in their own token based on how much volume they generated. LooksRare launched with a program of exactly this kind, distributing a daily pool of LOOKS tokens in proportion to each trader’s share of platform volume. The LooksRare documentation sets out the formula and notes that the program was discontinued in September 2023 in favor of a different approach.
The problem with paying for volume is that volume is trivial to manufacture. If the tokens you earn from a round trip are worth more than the fees you pay to make it, the trade is profitable no matter what the NFT is worth. That turns the reward program into a machine for printing fake activity. The same logic applies wherever trading volume decides an airdrop allocation, a leaderboard position, or a fee rebate.
Pushing a floor price before a real sale
The second motive aims at a buyer rather than a marketplace. A holder sells a token to themselves at a high price, then does it again, and the collection’s sales history now shows a cluster of expensive trades. To someone checking recent sales before making an offer, that cluster reads as demand. The wash trader then lists the real token at a price anchored to their own fake sales.
Chainalysis studied this pattern by tracing sales into what it called self-financed addresses, meaning addresses funded by the seller or by whoever funded the seller. Its research on NFT wash trading identified 262 users who had sold an NFT to a self-financed address more than 25 times. One of them had done it 830 times.
The profit picture was uneven. Of those 262, 110 made a combined $8,875,315, while 152 lost a combined $416,984 once gas was counted. Most wash traders in the sample were not making money. The minority who were made a lot, and they made it from people who read the fake sales as real ones.
How common it is
Estimates vary because every estimate depends on which heuristics you trust, but the range is wide enough that the direction is not in doubt.
The Research Policy study by Brett Hemenway Falk, Gerry Tsoukalas and Niuniu Zhang applied machine learning to 42,442 OpenSea collections, 9,474 LooksRare collections and 7,495 Blur collections. As summarised by Boston University, it found suspicious patterns in roughly 38 percent of trades and 60 percent of traded value, with rates of about 31 percent on OpenSea and about 95 percent on LooksRare.
Independent on-chain analysis has landed in the same territory on value. The open-source filter maintained by the analyst hildobby, set out in Dune’s write-up, flagged more than $30 billion of Ethereum NFT volume as wash trading, close to 45 percent of the total. LooksRare and X2Y2 were the worst affected on that measure, at 98 percent and 87 percent of volume.
Worth noting where the two studies disagree. They land close on value, at 60 percent and 45 percent, but far apart on how many individual trades are involved, at 38 percent against roughly 1.5 percent. That gap is a methodology difference rather than a contradiction. Counting by trade depends heavily on how aggressive your heuristics are, while counting by value is dominated by a small number of very large fake sales that almost any filter catches.
Two things follow. The first is that a headline volume figure on its own tells you very little. The second is that the problem is concentrated rather than uniform. Marketplaces that paid for volume have far higher rates than ones that never did, and within any marketplace a handful of collections account for most of it.
Four patterns you can check by hand
Every one of these is visible in the activity or sales history tab of a collection page. You do not need a subscription or a query engine.
The same two wallets, over and over
Open the activity feed and read the buyer and seller columns rather than the prices. Genuine trading produces a long tail of addresses that mostly appear once. Wash trading produces short addresses that alternate. If the same pair shows up trading the same token in both directions, you are looking at a round trip, and a round trip has no legitimate reason to exist.
Three or more wallets passing an item in a loop is the same trick with an extra step, added specifically to defeat the two-wallet check.
A price that ignores the rest of the collection
Compare each recent sale against the collection floor and against the spread of sales over the past month. Real markets produce a distribution. One token going for forty times the floor while everything around it sits flat is not a market signal, it is a single trade with nothing behind it.
Be particularly careful when an outlier sale is recent and the token is now listed. That sequence is the setup described above, in order.
Volume that arrives without new owners
This is the most reliable check available on a public collection page, because it compares two numbers that must move together in a real market. Genuine buying raises volume and raises the number of unique holders. Wash trading raises volume and leaves the holder count where it was, because the tokens are circling inside a closed group.
If volume jumps and the owner count is flat or falling, the money is going in a circle. Our guide to how you know what an NFT is worth goes further into which collection metrics hold up under pressure.
Wallets funded from the same place
This one takes an extra click but it settles most cases. Take the buyer address and the seller address into a block explorer and look at where each wallet got its first funds. Independent traders have unrelated funding histories. A wash trading cluster usually traces back to one wallet or one exchange withdrawal, because somebody had to pay the gas for all of them.
Tools that filter it for you
Several data providers publish volume figures with wash trades stripped out, and reading one of those alongside the raw number on the marketplace is a fast sanity check.
CryptoSlam runs a filter across the collections it tracks and reports adjusted volume. Its exact rules are proprietary, which is a real limitation, and when the company widened its definition to cover certain market-making patterns it removed about half of Blur’s reported sales revenue. Blur’s team disputed that change on the grounds that the methodology could not be inspected.
The hildobby dashboard on Dune takes the opposite approach. The logic is published, so you can read exactly what it flags: sales where buyer and seller are the same address, back and forth trades between two addresses, addresses that have bought the same token three or more times, and buyer and seller wallets funded from a common source.
Neither is an oracle. A filter that is too loose misses coordinated groups, and one that is too tight strips out legitimate market making. Use them to check whether a collection’s adjusted volume is close to its raw volume. When the two numbers are far apart, that gap is the finding.
What to do with this before you buy
Treat volume as a claim that needs support rather than as evidence. Before committing funds, check that the holder count moved with the volume, that recent sales cluster rather than scatter, and that the wallets behind the largest sales are not obviously related.
Wash trading also raises your costs even when you are not the target. Inflated recent sales push up what sellers ask, and marketplace and network fees are charged on the price you actually pay. Our breakdown of how NFT marketplace fees work covers what those costs come to on a typical trade.
Finally, fake volume rarely travels alone. A collection with manufactured trading history often has other things wrong with it, and the checks in our guide to spotting a fake or scam NFT project are worth running at the same time.
Common questions
Is NFT wash trading illegal?
Wash trading is prohibited in regulated securities and commodities markets. Whether a specific NFT trade falls under those rules depends on the jurisdiction and on how the asset is classified, which is unsettled in most places. Separately, using fake sales to induce someone to buy at an inflated price can amount to fraud regardless of how the asset itself is treated.
Can a marketplace stop wash trading?
Not completely. A marketplace can remove the incentive by not paying rewards for volume, and it can filter suspicious trades out of the figures it publishes. Neither prevents someone from sending a token between their own wallets, because that is an ordinary transaction the chain has no reason to reject.
Does wash trading affect the floor price?
Not directly. The floor is the lowest active listing, and a wash trade is a completed sale rather than a listing. The effect is indirect. Fake sales at high prices change what sellers think their tokens are worth, and they raise their asking prices accordingly.
The one-line summary
Volume can be manufactured for the cost of gas, but ownership cannot, so judge a collection by how many separate people hold it and how its sales are distributed rather than by the number at the top of the page.
Leave a Reply