In computer science, conservative two-phase locking (C2PL) is a locking method used in DBMS and relational databases.

In contrast to Non-Strict 2PL, Conservative 2PL prevents deadlocks.

The difference between 2PL and C2PL is that C2PL's transactions obtains all the locks it'll need before it begins. This is to ensure that a transaction that already holds some locks will not block waiting for other locks.

In heavy lock contention, C2PL reduces the time locks are held on average, relative to 2PL and Strict 2PL, because transactions that hold locks are never blocked.

In light lock contention, C2PL holds more locks than is necessary, because it is hard to tell what locks will be needed in the future, thus leads to higher overhead.

Also, a transaction has to release all of its locks, and obtain all of them again, if it fails to obtain even one lock that it needs in the first run. Furthermore, each transaction needs to declare its read and write set (data items to be read/written during transaction), which is not always possible. Because of these reasons, C2PL is not used frequently in professional use.