Oxidative DNA damage detected by the comet assay as formamidopyrimidine DNA glycosylase (FPG) sensitive sites, almost as a rule is reported as comet assay score rather than numerical sites in the genome, probably because the latter requires X-ray calibration. We compared the ability of five experienced and five inexperienced comet assay investigators to detect a dose-response relationship in irradiated A549 lung epithelial cell culture samples (0, 10 Gy and three samples of 5 Gy), based on an arbitrary five class scoring system. The samples were scored on three different occasions, thus allowing determination of the variation in sample scoring. All investigators qualitatively distinguished between samples in a dose-dependent manner, albeit with large variation in the slope and intercept of dose-response curves. There was a tendency that investigators with experience in scoring A549 cells had more consistent results than experienced investigators who had only scored lymphocytes or inexperienced investigators. The inexperienced investigators improved their scoring ability during the three sessions. Subsequently we showed that the variation in baseline level of FPG modifications in mononuclear blood cells of five healthy humans was lower when investigators used their individual X-ray calibration curve as compared to a common calibration curve. In conclusion, this study showed that comet assay investigators score differently when using a five class scoring system, which indicates that more consistent estimations of FPG sites in the genome are obtained by use of investigators' individual X-ray calibrations.