The traditional way of assessing students’ oral performance is by way of formal oral tests, administered at the end of a teaching period. Typically, students sit in pairs in front of the examiner, many times their own teacher, other times another teacher for an “unbiased judgment”, and perform a task based on a given situation. For test security, reliability and, arguably, validity purposes, the students cannot know in advance the task they are going to be asked to perform. The teacher listens, takes notes, and gives a grade based on a set of scoring rubrics, and students have only one shot at the task. This is the model used in high stakes proficiency tests and replicated in EFL/ESL classes due to the belief that our role is to prepare students for standardized proficiency tests. Or is it?