This article argues that overt multiple wh -fronting in languages like Bulgarian consists of moving a single wh -cluster to [Spec, CP]. The formation of wh -clusters is motivated by the assumption that wh -elements can act as landing sites for wh -movement due to morphological properties of wh -words. I further argue that languages such as Japanese constitute covert instances of this process of wh -cluster formation, accounting for intricate constraints on multiple wh -questions such as the so-called “additional- wh effect.” Another central claim of the article is that despite appearances, multiple wh -questions in German equally involve the formation of wh -clusters, which are shown to consist of one visible and one or more invisible wh -elements. This analysis provides a new account for the lack of “short” and the presence of “long” superiority effects in German.