Should test methods be documented?
If the method name is expressive, we think there is usually no need to comment it. But as with any other method, documenting them gives you time to re-think about what you are testing and it also enables you to pass this information on to other developers (or to yourself a couple of weeks later).
In Growing Object-Oriented Software, Guided by Tests, Steve Freeman and Nat Pryce advice to name tests in terms of the features that the unit under test provides. Since test methods are only called by the test framework through reflection, we can give test method names any length. E.g. (taken from their book) for the features of a list:
This convention has been invented by Chris Stevenson. You may have a look at the test dox report of Stevenson's TestDox project and its unit test report to get further examples for this approach.