Abstract
The study of regulation has previously involved indirect, and largely correlative, approaches. In the following contribution I illustrate the uses and limitations of these traditional approaches, and then discuss how molecular genetics provide a new tool to test directly ideas about regulation in vivo in the intact organism.