I remember being quite amused sometime last year (before I was a blogger) when coming across the following one-page paper on Noam Elkies’ page. The story is that the paper was rejected from the Monthly problems section because on referee thought it trivial and the other too difficult. I think the problem itself (an application of Zorn’s lemma) speaks for itself.
Incidentally, the argument shows that the same is true if “measure zero” is replaced with “of the first category.” I’m curious about other examples of this phenomenon. (Stated axiomatically, it becomes trivial though—just a restatement of Zorn’s lemma.)