This paper examines the formation of passive verbs and its distribution and productivity in Modern Standard Arabic and Palestinian Arabic (hereafter MSA and PA, respectively). I will account for the differences between the morphology of passivization in both languages, arguing that it results from the component of the grammar, the lexicon or the syntax, where the formation takes place. Passivization in MSA is highly productive, almost exception free and demonsrates transpatent morphological relations between passive verbs and their active counterparts. PA passivization, in contrast, is less productive, its morphology is less transparent and it is subject to morphological constraints that block the derivation of many forms. In addition, I will show that a word-based derivation captures the generalizations about passive formation in these languages and therefore should be favored. Blocking effects on passive formation can only be explained if the mechanism that forms words relies directly on existing forms rather than on a consonantal root.