In Chapter 1 , I elaborated a principal-agent model of delegation to, and agency and agenda setting by, supranational actors. In this chapter, I turn to the empirical study of delegation and discretion to the European Commission, examining both the functions delegated to the Commission and the patterns of delegation and discretion allocated to the Commission in primary and secondary EU legislation. Specifically, I utilize both qualitative and quantitative data to test two hypotheses specified in Chapter 1 . First, my principal-agent model predicts that member states will delegate certain types of functions to supranational agents, namely, monitoring compliance, filling in incomplete contracts, setting the legislative agenda, and providing expert and credible regulation. In this chapter, I therefore analyse the functions delegated to the Commission in both primary (treaty) and secondary legislation, which I argue fit closely the functions predicted by the theory. Second, I test the hypotheses derived from Epstein and O'Halloran (1994 ; 1999a , b), Huber and Shipan (2000) and others, that agency discretion—defined as a function of the control mechanisms set down in an act of delegation—varies systematically across issue areas as a function of the demand for (1) expert information and (2) credible commitments. I also test Majone's (2001) hypothesis that delegation in the treaties is motivated primarily by a desire to secure credible commitments, and therefore features extensive discretion to EU organizations; while delegation in secondary legislation is motivated primarily by informational concerns, and is accordingly accompanied by the more extensive use of control mechanisms and thus lower levels of supranational discretion.