This study examines the efficacy of user participation in developing an accounting application. The research takes place over a 19-month time frame, involves 516 clerical-level accounting subjects, and includes experimental manipulations in a field setting. The model of user participation and involvement proposed by Hartwick and Barki (1994) provides the foundation for the research framework. Their model is augmented by the inclusion of concepts from procedural justice and self-efficacy research. Participation is manipulated at three increasing levels: (1) no voice, (2) non-instrumental voice, and (3) instrumental voice. Research findings suggest that users' pre-experiment level of involvement with and attitude toward the present system are positively associated with their desire to participate in the development of the new system. Study results also indicate that users' a priori self-efficacy beliefs regarding their perceived ability to effectively contribute to the development process are positively related to desired participation. Pre- to post-experiment gains in psychological and behavioral variables are next assessed. In the instrumental voice condition, user involvement, user attitude, and performance gains are significantly highest. User attitude and involvement gains are significantly higher in the non-instrumental voice condition than in the no voice condition; however, gains in user performance are not significantly different between these treatment conditions. Research findings indicate that user participation can be effective, particularly when users perceive a noticeable degree of instrumental control over the decision outcome.