A transaction identifier is calculated from a transaction's encoded data according to its network's rules. It lets software refer to a specific transaction without repeating its complete contents. In Bitcoin, inputs identify a previous output using the previous transaction's identifier together with an output index. The identifier therefore helps connect the graph of spending activity. It should always be interpreted alongside the network because explorers, test networks, and different chains have separate histories.

Suppose a sender provides an identifier after initiating a transfer. Searching an appropriate explorer may reveal an unconfirmed transaction, an included transaction, or no known record. Inclusion still leaves further checks: the recipient, asset contract, amount, execution status, and relevant confirmations. A transaction can exist while transferring a different token with a familiar name, or while failing during contract execution. The identifier is a pointer to evidence, not the conclusion that the expected payment arrived.

Segregated Witness distinguishes a Bitcoin transaction's txid from its witness transaction identifier, or wtxid. The latter includes witness data; the former excludes it under the defined serialization. This distinction helps address transaction malleability and supports witness commitments, but software must still use the appropriate identifier for each protocol operation. A wallet replacement can also be a genuinely different transaction with a different hash. Record the final accepted transaction instead of assuming the first displayed identifier is permanent.