Indirect system use refers to a user (the principal) performs IS-related tasks via another user (the agent). Despite the prevalence of indirect system use, the extant literature on system usage and its performance mainly focuses on direct system use. During the process of indirect system use, the goal conflict and information asymmetry could arise between the principal and the agent, which create challenges of appropriately managing the agent’s behavior. Drawing on the Agency theory, we propose that indirect system use can be categorized into two types: behavior-oriented indirect system use and outcome-oriented indirect system use. A conceptual model is constructed to theoretically understand how these two types of indirect system use impact task performance differently and contingent on the extent of task complexity.