Practicing user interface designers and software developers need more than just rules and guidelines, they need practical tools and methods that point the way toward simpler and more supportive systems by highlighting the essentials of user intent and interaction. User interface design often focuses on the rules and the technology for the system interface itself, while techniques such as work-flow modeling and contextual inquiry 193 have concentrated on the work and tasks. What is needed is a bridge that helps practitioners directly connect the structure of the user interface to the structure of use. Persistent questions are at the heart of this need. How do we distinguish what is truly nec-essary to support the work from what users may say they want or from what we are by habit prepared to offer? How can we describe and represent this work, and how can we then use our understanding of the work to design the architecture of the user interface to support it? ssential use case modeling offers an approach to these questions that can be used within almost any usercentered design strategy it was developed over several years by Lucy AD Lockwood and the author [G, 8] as a process that would be easy to learn and to apply, yet would dependably point toward user interfaces that fit markedly better with intended uses. This process is woven from two conceptual threads in systems analysis and design. The first thread is the concept of use cases, originally devised by Ivar Jacobson as a tooi for object-oriented so&vare engineering [IO]. Use cases represent what a system offers to its users, the functionality of the system as viewed from the outside. When use cases are combined with the second thread, essential modeling, the result is a process that smoothly connects the design of user interface architecture back to the essential purposes of a system and the work it supports. This connection is achieved through three interdependent models: the user role model, the essential use case model, and the use context model.