Usual derivations of Lilders's projection rule show that Liuders's rule is the rule required by quantum statistics to calculate the final state after an ideal (minimally disturbing) measurement. These derivations are at best inconclusive, however, when it comes to interpreting Liuders's rule as a description of individual state transformations. In this paper, I show a natural way of deriving Liiders's rule from well-motivated and explicit physical assumptions referring to individual systems. This requires, however, the introduction of a concept of individual (...) state which is not standard. (shrink)