Abstract: In prior works, Corning showed a significant variation of burn results of an identical fiber optic cable tested in different test facilities. The purpose of this study was to identify and understand the sources of those deviations. A calibration cable was established, possessing a low complex structure which decreases the variances originated from the cable itself. Prior to understanding deviations of results, the reliability of the equipment itself was determined over a significant time period using the calibration cable. The so-gained confidence level using the calibration cable is used to judge the deviations from different laboratories. Moreover, the test and the procedures according to EN 50399 are examined related to sensitivity of results to variations of the test parameters. In addition, modeling was used to understand effects and predict the behavior. The test itself, according to EN 50399, is repeatable within its confidence level at one facility. Possible sources for existing deviations between different labs are examined and discussed.