This book proposes a new solution to the long-standing puzzle of auxiliary selection in Romance languages, in particular Italian. The following questions are addressed: why the perfect auxiliary appears in the two forms be and have within a single language, what drives this distribution, and how cross-linguistic data can be accounted for. The solution to these issues consists of an Agreebased analysis that accounts for auxiliary selection in root clauses and restructuring in Standard Italian and in Italo-Romance varieties, which is also compatible with participle agreement. By answering these questions, the book also touches upon more theoretical and foundational problems, such as the distribution of labor between syntax, morphology and the lexicon, and the conditions on the operation Agree (in particular, multiple probing, locality, and minimality). This work contributes to the discussion in the fields of formal morpho-syntax, theoretical linguistics, and Romance linguistics.