This paper discusses replication studies in the context of computer-assisted language learning. I begin with what is feasible for CALL replication studies, proposing that conceptual replication is more realistic than exact replication. Then, based on the premise that theories and principles of SLA should underlie CALL research, a brief summary of CALL's contributions to SLA is given. I attempt to set an agenda for CALL replication studies, but rather than suggesting a list of specific studies that could or should be replicated, I propose an expansive approach and provide guidelines for how to select studies or types of studies that could be replicated conceptually. This approach allows researchers to select from the most pressing issues in SLA/CALL studies and conduct conceptual replication, based on their theoretical perspective and on a principled implementation of the ever-evolving new technologies available. I close with some examples of concepts or notions that need to be tested or verified across different contexts, including my own pet areas of interest, and how one might go about replication studies in these areas.