The present study investigated differences in visual expertise across levels of proficiency in train traffic control during a simulated scenario. Eye tracking metrics found to correlate with expertise reported in a meta-analysis on visual expertise were used. The aim of the study was to investigate whether the same results found in the meta-study could be obtained in the less controlled and dynamic work environment of train traffic control. Studies of this character are rare and also notoriously difficult to conduct due to a high level of potential noise. Results of the study indicates that eye behaviour seemed to correlate with years of experience also in a more naturalistic setting, but it did not correlate with expert ranking by instructors or a post-hoc measure of proactivity in task performance. A discussion is provided where a delineation of experience and expertise is made in light of differences between eye movement behaviour and cognitive aspects of problem-solving.