David at E.G. suggests a potential objection to consequentialism. Suppose that a painting, based on stolen photos, is considered a great work of art because of its immoral origin. That is, its aesthetic value depends upon the immorality of its creator. Great aesthetic value outweights the disvalue of petty theft, so according to consequentialism the creation of the painting was not immoral after all. But we have assumed the immorality to be an essential component of the painting's value; if it were created in a moral fashion, it would lose this value. Absent this resulting value, the theft is immoral. (So the value returns, and we spin round the circle again.) We are left with a contradiction: the painter is immoral iff he is not. This is impossible, so one of our assumptions must be false.
David suggests we hold consequentalism to blame. But I think this is a more general problem to do with second-order values. For imagine the following: Nasty Adolf desires that more desires be thwarted than fulfilled. Further suppose that - apart from this desire of Adolf's - there are in total an equal number of fulfilled and thwarted desires in the world. It follows that Adolf's desire is fulfilled iff it is not fulfilled. Again we have a contradiction, but this time without any mention of consequentialism. [Update: see also my post on contextual impossibility.]
So what is the real problem? It seems a lot like the liar paradox. A more direct analogue might be "I wish that this wish won't come true", or equivalently, "I desire that this desire be thwarted". But it should be clear that Adolf's desire is problematic in the same self-referential way, though by a more indirect route. The same is true of our painting's aesthetic value. (We might re-describe the situation as follows: The collectors are desiring that the painter thwarts more desires than he fulfils, where this is the deciding desire!)
It seems we must conclude that such desires and values are impossible, much as the statement that "this statement is false" is impossible. (A little voice objects: "But you just made that statement, so it mustn't be impossible after all!" But no, all I did was write some words, there's no guarantee that the sentence had any meaning. If it is just so much gobbledegook masquerading as English, then it does not in fact state anything at all.) This response is awfully ad hoc, of course. It sure seems possible to make such a statement, or for such desires and values to exist. But denying this possibility is one (possible) solution, at least.
Anyway, my main point is just that David's scenario reduces to the value-based variant of the liar paradox. The problem can be restated without appeal to consequentialism at all. So it cannot work as an objection to consequentialism. The problem lies elsewhere.