The diversity of the answers given so far clearly illustrate the vagueness of the original question.
For a precise answer you need to specify along which dimension(s) you wish to measure textual similarity. The techniques to recommend fully depend on what you want to measure. Possible textual features to consider are:
Text length
Text formatting
Words and/or n-grams used
Stylistic aspects
Topics covered
… to name but a few.
The suitable similarity measure you are asking for depends on which dimensions you are interested in. Cosine similarity works well for 3 while LSA is good for 5, etc.