This research extends extant literature examining the relationship between IS use and performance. While prior theory has predominantly treated IS use as an antecedent of performance and hypothesized a positive effect of use on performance, this research provides an alternative perspective. Specifically, this paper theorizes that under certain contexts performance can be an antecedent of use and that the effect can be in the opposite direction. In contrast to non-contingent models, this paper proposes a contingent model in which the IS use-performance relationship is contingent on organizational performance and varies over time. The model proposed here is tested on longitudinal data. Distributed lag model was employed for data analysis. The results support our hypotheses that performance is an antecedent of IS use and that the effect is negative, i.e. when organizational performance declines, IS use increases but after a period of increased use, the effect declines to non-significant levels.