In this paper I argue that the acceptance of an absolutely unrestricted quantification implies the existence of an absolutely empty possible world. This result could be relevant because David Lewis both admits an absolutely unrestricted quantification (for example in Parts of Classes) and rejects the existence of an absolutely empty possible world (in On the Plurality of Worlds). In order to vindicate my thesis, I propose two strategies. The first is based on the assumption that the phrase ‘nothing’ cannot be always reduced to a quantifier phrase, as Graham Priest and Alex Oliver with Timothy Smiley have argued. This strategy consists in a paraphrase of the notion of everything that constrains us to admit an empty possible world. The second strategy mainly consists in the use of an “idealistic” principle (say «every determination is negation») and its consequences.