Why a router compares multiple fee tiers

A fee tier identifies one part of a pool's pricing terms. It does not describe the pool's full executable output. Separate tiers can contain different liquidity, so a router must quote the requested amount against each eligible pool.

Different tiers are different markets

In a concentrated-liquidity design, the same token pair can have several pool configurations. Providers choose where to place liquidity. The pool with a lower fee can consequently have less useful liquidity along the trade's path.

Uniswap's v3 introduction describes multiple fee tiers as separate liquidity opportunities. Its path specification includes fee information when identifying a pool.

The router compares outcomes

Imagine two hypothetical A/B pools. Pool L has a smaller percentage fee but very little active liquidity near the current price. Pool H charges more per unit of input but has much greater capacity in the relevant range. Either can win for a particular amount; the fee label alone cannot decide.

The router can also allocate different portions to the two pools. That allocation treats them as independent sources only if their states are genuinely separate. A single protocol logo in the interface may therefore conceal several fee-tier markets.

Why the selected tier can change

Liquidity additions, withdrawals and trades can change the relative attractiveness of the tiers. Changing the input amount can also alter the result because curve effects grow differently across the pools.

When a route selects a higher-fee tier, the useful question is whether the complete route provides stronger terms under its objective. It is not enough to compare the posted fee percentages in isolation.

This also explains why “supported fee tiers” and “used fee tier” are different facts. An application can support several and choose only one for the current order. The route's pool identity tells you which state and rules the quote actually uses.