Which Roblox limiteds actually sell? 90 days, measured

Every limited has a RAP, so every limited looks like it has a price. Far fewer have buyers. We wanted to know how many, so we counted. Our limiteds file, refreshed from Roblox's own resale data on October 27, 2026, holds 172 classic limiteds, every one of them created by Roblox. In the 90 days before that refresh they sold 194,335 units between them, with the newest sale in the file dated October 26.

Spread evenly, that would be about 1,130 sales per item. It is nothing like even. The median item sold 233 copies in the 90 days, 46 items sold more than a thousand each, and 18 sold none at all. Most of the trading in this market happens on a short list of cheap items, and the most expensive items hardly trade.

How we counted a sale

Roblox's resale data gives each item a daily series: how many units sold that day and the average price they sold at. Our builder adds up the units for the last 90 days. It skips any day whose price sits under a tenth of the item's RAP or over ten times it, which it treats as a glitch or troll trade, and a skipped day's units are not counted either. A unit is one copy changing hands.

That is all the data can say. It does not list individual sales, so we cannot see who bought or sold, how many different people traded, or what anyone paid outside the daily average. It also says nothing about listings: an item can have dozens of copies for sale and no buyer, and this count would show zero.

Five groups, very uneven

We sorted the 172 items by how many copies each sold in the 90 days.

The 46 items that sold 1,000 or more copies are 26.7% of the list and account for 86.6% of the units sold, 168,301 of them. Below them, 67 items sold between 100 and 999 copies, together 12.8% of the units. Everything else is thin. The 59 items that sold fewer than 100 copies, about a third of the list, sold 1,142 units between them, 0.6% of the total. Thirteen of those sold between 1 and 9 copies in three months, 42 units in all.

Put another way, 113 items sold at least 90 copies, an average of one a day or more, and 47 sold at least 900, ten a day or more. The other 59 sell less than once a day on average, and many far less often.

The dates in the file point the same way. 112 items recorded a sale on October 26, the newest day in the file. 141 had sold within the week before the refresh, 152 within 30 days and 154 within the 90-day window. Of the 18 that sold nothing, four last sold within the past year, nine last sold more than a year ago, and five have no sale on record that passes our glitch filter at all.

The fifteen that do the trading

Rank the items by units sold and the top of the list looks nothing like the top of a value list.

8-Bit Royal Crown sold 8,848 copies, more than any other item, and on its own that is 4.6% of everything sold. The top five made 18.8% of all units, the top ten 32.8% and the top fifteen 44.6%. Eighteen items account for more than half.

Thirteen of the fifteen have a RAP between 172 and 350 Robux. Six have the word "Egg" in their names, and a seventh is Eggraging Shark of the Sea. The two exceptions are Black Iron Branches at 1,292 and 8-Bit Royal Crown at 28,271, the only item in the top fifteen with a five-figure RAP. That combination of a mid-range price and heavy trading is why the crown also made the largest single share of the fee Roblox kept in our October 23 measurement of the 30% cut.

Heavy trading by count does not mean a large share of copies changes hands. Roblox's catalog reports how many copies of each item exist, and the cheap favourites exist in the millions. Cake Topper sold 7,476 copies in 90 days, out of 4,395,926: 0.17% of them. 8-Bit Royal Crown's 8,848 sales came from 251,515 copies, 3.5%. The highest share on the list belongs to Infernal Undead Immortal Sword, RAP 20,692, which sold 589 of its 12,693 copies, 4.6%. The median item sold 0.54% of its copies in the window.

Price decides who trades

The clearest pattern in the file is between price and volume. Ranked by RAP and ranked by units sold, the 172 items line up almost in reverse: the rank correlation between the two is -0.86, where -1 would mean the two rankings are exact mirror images.

The bands show the slope directly.

Every item under 100K Robux sold at least once in the 90 days. Above that line, sales thin out fast. The 24 items in the 100K to 1M band have a median of 25.5 copies sold (the middle two sold 22 and 29), the typical item between 1M and 10M sold one, and none of the four above 10M sold any. Fifteen of the 18 items that sold nothing have a RAP of a million or more, and the median RAP of the 18 is 4,604,224 Robux.

The quiet end is a list of famous names. Dominus Frigidus, the highest RAP in the file at 28,059,680, last sold on January 19, 2023, by the dates in Roblox's data. Dominus Astra and Dominus Empyreus, both above 13 million, have no sale on record that passes our filter. Dominus Infernus and Eerie Pumpkin Head have no sale history in the data at all. Their RAPs are real numbers from Roblox, but none of them rests on a sale in the last three months.

What this means if you are buying or trading

A RAP is an average of past sales, so its meaning depends on how many sales sit behind it and how recent they are. On an item that sells a thousand copies a quarter, RAP is close to what the next buyer will pay. On an item that sold one copy, or none, it is a memory. We covered the theory in a high RAP with no sales is a trap; the counts above show how much of the market that warning covers. About a third of the classic limiteds we track sold fewer than 100 copies in three months.

Our limiteds market page prints this as an activity label beside each item's RAP, from the same 90-day count: Dormant (none), Low (under 50), Moderate (under 500), Active (under 5,000) and Hot. On October 27, 2026, 18 items were Dormant, 31 Low, 60 Moderate, 54 Active and 9 Hot. If you track limiteds, the label is worth reading before the price.

The same applies to trades. The limiteds trade checker compares the value estimates on both sides, and says on its own page that how fast an item re-trades does not enter the calculation. Two sides can total the same and still differ in how easily each could be sold again: a side made of Hot items, which average dozens of sales a day, is easier to sell on than a side built on one Dormant item that has not found a buyer in three months.

People also search for limited "snipers", tools that try to buy a copy listed below its usual price. Our data shows sales, not listings, so it cannot say where an underpriced listing exists, and we do not cover sniping tools. What it does show is where a usual price exists at all: on the 59 items that sold fewer than 100 copies in 90 days, there is too little trading to say what an item usually sells for.

What these numbers can and cannot tell you

The window is the 90 days before October 27, 2026. A longer window would put more sales on the rare items, and a single sale moves an item from Dormant to Low without anything changing in the market. The counts are daily totals, so a day with a glitch-priced average drops out whole, even if real sales happened on it as well.

The 172 items are the classic Roblox-created limiteds our file prices. The modern collectibles we list without a price and every community-created limited are not in these figures. The copy counts come from Roblox's catalog and say how many copies exist, not how many are held by active players.

The file every number came from is saved with the site, at /data/snapshots/2026-10-27/limiteds.json, so each figure here can be checked against the exact data it used.

Liquidity is concentrated

A limited's price and its saleability are two different facts, and on this list they run in opposite directions. The cheap end trades constantly, on items that exist by the million. The expensive end holds the biggest RAPs and the fewest buyers. Before treating any limited as worth its RAP, check how many copies sold in the last 90 days and when the last one went.