Prediction has long been considered advantageous in simultaneous interpreting, as it may allow interpreters to
comprehend more rapidly and focus on their own production. However, evidence of prediction in simultaneous interpreting to date is
relatively limited. In addition, it is unclear whether training in simultaneous interpreting influences predictive processing
during a simultaneous interpreting task. We report on a longitudinal eye-tracking study which measured the timing and extent of
prediction in students before and after two semesters of training in simultaneous interpreting. The students simultaneously
interpreted sentences containing a highly predictable word as they viewed a screen containing four pictures, one of which depicted
a highly predictable object. They made predictive eye movements to the highly predictable object both before and after their
training in simultaneous interpreting. However, we did not find evidence that training influenced the timing or the magnitude of
their prediction.