This paper identifies and examines the logical relationships commonly found in computer-assisted telephone interviewing (CATI) instruments and describes a method of programming a computer to recognize these relationships as logical structures. CATI systems can use instrument logic databases to generate instrument flow charts, display instrument logic at interviewing stations, and augment or replace keyboard-based interview movement commands with mouse-based commands. The paper includes a discussion of how the properties of logical structures can be applied to give the instrument designer more control over the instrument's content and function.