The goals of this talk are two-fold. Its primary aim is to provide a proper treatment of habituals within the framework of event semantics, which would (a) allow a unified analysis of the habitual and the progressive as varieties of the imperfective, (b) work both for "bare habituals" and for "quantified habituals", and (c) model the scope behavior of indefinites in habituals of both types. In addition, the talk will also show that an analysis of habituals that meets these desiderata can't be implemented in combination with a classic GQ analysis of quantificational NPs and, thus, will provide a further argument in favor of event-based approaches to quantification.