This study experimentally investigated the effectiveness of feedback on learner refusals in a computer-simulated academic advising session. Ninety participants were assigned to one of three conditions: implicit feedback, explicit feedback, and comparison group. Oral and written discourse completion tasks (DCTs) were administered in a pretest immediate posttest design. Pragmatic development was identified by examining uptake, which was operationalized as incorporation of pragmatic features from the simulation that were absent on the pretests. This analysis, therefore, focuses on the subset of learners who did not make use of the target pragmatic features from the simulation on either pretest (n = 59). Participants in the explicit feedback group showed a significantly greater degree of uptake on the posttest ODCT than the comparison group, while both feedback groups showed significantly greater uptake than the comparison group on the posttest WDCT. The study also used a written retrospective comparison task to examine how learners described changes in their language use following treatment. Responses on this task indicated a greater degree of attention to (im)politeness and evidence of learning among participants in the feedback groups, while participants in the comparison group primarily described their responses as the same as before treatment.