Clinical practice guidelines have enormous potential to improve the quality of and accountability in health care. Making the most of this potential should become easier as guideline developers integrate guidelines within information systems and electronic medical records. A major barrier to such integration is the lack of computing infrastructure in many clinical settings. To successfully implement guidelines in information systems, developers must create more specific recommendations than those that have been required for traditional guidelines. Using reusable software components to create guidelines can make the development of protocols faster and less expensive. In addition, using decision models to produce guidelines enables developers to structure guideline problems systematically, to prioritize information acquisition, to develop site-specific guidelines, and to evaluate the cost-effectiveness of the explicit incorporation of patient preferences into guideline recommendations. Ongoing research provides a foundation for the use of guideline development tools that can help developers tailor guidelines appropriately to their practice settings. This article explores how medical informatics can help clinicians find, use, and create practice guidelines.