CS2 Trade Up Contracts Explained
A trade up contract consumes ten skins of the same rarity and returns one skin of the next rarity up. Unlike case openings, the mechanics are fully deterministic in the parts that matter — which means a trade up can be calculated in advance, and occasionally calculated into a profit.
The rules
- All ten inputs must be the same rarity. You cannot mix a Restricted in with nine Mil-Specs.
- All ten must be either normal or all StatTrak™. The output matches the inputs.
- The output is one rarity higher, drawn from the collections your inputs came from.
- Covert is the ceiling. You cannot trade up into knives or gloves.
- Souvenir items cannot be used.
How the outcome is chosen
This is the part people get wrong. The game does not pick a collection and then a skin — it weights by input. Each input skin contributes one entry, and that entry points at the pool of next-tier skins in its own collection.
So if you use seven skins from Collection A and three from Collection B, you have a 70% chance of landing in A's next-tier pool and 30% in B's. Within the chosen collection, each eligible skin is equally likely.
The immediate implication: a collection with only one skin at the target rarity is far more predictable than one with five. Serious trade up planning starts by finding collections with narrow output pools.
The float formula
The output's wear is not random. It is derived from the inputs:
Output float = (OutputMax − OutputMin) × AverageInputFloat + OutputMin
Where AverageInputFloat is the plain arithmetic mean of your ten input floats, and OutputMin / OutputMax are the float caps of the specific skin you receive.
Two things follow from this. First, low-float inputs produce low-float outputs — which is why people hunt for the cheapest possible Factory New fodder. Second, the same average produces a different wear on every possible output, because each output skin has its own caps. A 0.05 average might give you a Factory New from one skin and a Minimal Wear from another. Every skin page on this site lists that skin's float cap, which is exactly the number this formula needs.
When trade ups actually profit
Sometimes they do. The conditions are narrow and all of them have to hold at once:
- The output pool is small and mostly valuable. One expensive outcome among five cheap ones rarely clears; three expensive outcomes among four is a different calculation.
- Input fodder is genuinely cheap. Your cost is ten items plus the effort of sourcing them. If the fodder itself has appreciated, the margin is gone.
- The float target lands in a valuable wear. Getting the right skin in Battle-Scarred can still be a loss. Run the formula before you buy the inputs, not after.
- You account for fees. If you plan to sell the output on the Steam Market, subtract the 15% cut before calling it profit. Trading the output directly avoids this entirely.
- You can absorb variance. A contract with a positive expected value still loses most individual attempts if the good outcome is one in five. Positive expected value only pays out over volume.
Community trade up calculators handle the arithmetic and are worth using. What they cannot do is tell you whether the prices they pulled are still accurate, or whether the output will actually sell at the quoted number.
The honest summary
Trade ups are the one item-sink in CS2 where careful players can find an edge, because the mechanics are public and the inputs are priced by a market that does not always reprice fodder when an output moves. That edge is small, it requires volume, and it disappears the moment you stop checking current prices.
For most people, the better use of a pile of low-tier skins is simpler: bundle them into a trade. Ten unwanted Mil-Specs that a calculator says are worth a few dollars are often worth more as a sweetener toward an item you actually want, with no variance at all.
Browse live listings to see what people accept in bundles, or list your fodder and find out.