Some people have supposed that utility is good in itself, non-in-strumentally good, as distinct from good because conducive to other good things. And in modern versions of this view, utility often means want-satisfaction, as distinct from pleasure or happiness. For your want that p to be satisfied, is it necessary that you know or believe that p, or sufficient merely that p is true? However that question is answered, there are problems with the view that want-satisfaction is a non-instrumental good. What if you want something only because you have a false belief? What if the time at which you want that p is fifty or five hundred years before the time to which p itself refers? To meet these difficulties, qualifications have to be introduced, and much has been written about how exactly these qualifications are to be framed.1 There is however what may be a rather more serious objection to the view that want-satisfaction is a non-instrumental good, or rather to the combination of that view with the principle that it is sufficient for your want that p to be satisfied simply that p is true. The objection is that this combination forces you to give undue weight to the mere acquisition of desires when you come to make judgements about changes in the value of things. It forces you to say that for any true proposition p, which initially you do not want to be true, your mere acquisition of a desire that p will, other things being equal, make the world better. Non-instrumental value can be increased merely by multiplying desires, even though everything else remains the same. Surely, however, improving the world is not as easy as that.