Theoretical explanations of the relationship between creativity and self-control most often suggest a directional relationship whereby self-control influences creative processes. Controlled experiments testing this assertion have provided conflicting findings, which may stem from systematic differences in the administration and scoring of the divergent thinking tasks employed. The current study was conducted to examine the effects of ego depletion on different components of divergent thinking as a function of task instruction. Participants (N = 152) were randomly assigned to one of six cvnditions, based on a Depletion (ego depletion vs. control vs. difficult, nondepletion) X Task Instruction (creativity vs. fluency) design. After completing a well-established ego-depletion procedure. participants completed three alternate uses tasks with instructions that prompted either fluency or creativity. Analyses of the responses (scored for fluency, statistical uniqueness, and subjectively rated creativity) revealed a significant interaction effect on the subjectively rated creativity of the responses, wherein ego-depleted participants exhibited poorer creativity when task instructions prompted creativity and greater creativity when task instructions prompted fluency. Although participants who received task instructions prompting creativity (as opposed to fluency) provided significantly more creative responses in the two control conditions, task instructions had no effect for those in the depletion condition. The main effect of task instruction on uniqueness scores was also significant (i.e., participants asked to be creative provided more unique responses). These findings shed light on the role of self-control in divergent thinking and have important implications for future studies.