In statistics, the Lehmann-Scheffé theorem states the any estimator that is complete, sufficient, and unbiased is the unique best unbiased estimator of its expectation.

The usual way to find such an estimator is application of the Rao-Blackwell theorem.