A conceptual framework is presented that resolves some of the conflicts and inconsistencies in the various paradigms pertaining to goal setting. Hypotheses about the impact of quota level, expectancy, and self-efficacy on motivation are developed and tested in a laboratory experiment in which the subjects assume the role of salespersons and negotiate with opponents whose roles are simulated by a custom-designed computer program. The results indicate that as quota level is increased, effort increases only up to a point, after which increases in the level of the quota may actually decrease effort. Additionally, the impact of increased quota levels is stronger for subjects who are high in self-efficacy than for subjects who are low in self-efficacy. Finally, information about the level of task difficulty also influences the motivation to expend effort at the task.