Joint application design (JAD) is an intuitively appealing systems development method that encompasses approaches for enhancing user participation, expediting development, and improving the quality of specifications. Although the JAD method has been widely acclaimed, little is actually known about its effectiveness in practice. This paper reports findings of a field study at three organizations in which JAD practices were examined to assess how practice influenced outcomes of JAD use. Findings suggest that the organizations realized modest improvements in systems development outcomes by using the JAD method. JAD use was most effective in small, clearly focused projects and less effective in large, complex projects. Findings also suggest that practices for using JAD at the research sites reduced its potential effectiveness by limiting user participation, focusing workshop exercises on developers' analytical models and terminology, and applying the JAD method prematurely in large projects. Adaptations of the JAD method in practice stemmed from unresolved conflicts between assumptions underlying the espoused JAD method and the status quo IS development approach at each organization. Implications for the effectiveness of the JAD method as a software process improvement are considered. (C) 1999 Elsevier Science Inc. All rights reserved.