Early researchers suggested that virtual humans (animated pedagogical agents) required a persona if they were to be anthropomorphized by the user. Over the past two decades, research has focused on design features that target the perception of persona in order to increase the anthropomorphization of the agent toward the end of increased learning outcomes. This systematic narrative review analyzes studies that have measured agent persona using the Agent Persona Instrument to assess whether there is evidence that persona increases learning outcomes. Our findings suggest that significantly higher persona ratings have little effect on learning outcomes. However, design features such as facial expressions and gestures were more important moderators of agent persona than voice and agent type, and environmental details such as study domain or student population suggested no clear direction in the perception of persona. The article concludes with a discussion of future directions for study designs and potential frontiers.