The comet assay is a simple technique for measurements of low levels of DNA damage and repair in single cells. However, there is variation in background levels of DNA damage in peripheral blood mononuclear cells (PBMCs). This variation has been documented by inter-laboratory ring-trials where identical samples have been analysed in different laboratories using the formamidopyrimidine DNA glycosylase (Fpg)-modified comet assay. The coefficient of variation of background levels of Fpg-sensitive sites was 128 % in the first inter-laboratory validation trial called European Standards Committee on Oxidative DNA Damage. The variation was reduced to 44 % by the end of the project. Subsequent ring-trials by the European Comet Assay Validation Group showed similar inter-laboratory variation in Fpg-sensitive sites in PBMCs (45 %). The lowest inter-laboratory variation in Fpg-sensitive sites in PBMCs was 12 % when using calibration to standardize comet assay descriptors. Introduction of standard comet assay procedures was surprisingly unsuccessful as certain laboratories experienced technical problems using unaccustomed assay conditions. This problem was alleviated by using flexible assay standard conditions rather than a standard protocol in a ring-trial by the hCOMET group. The approach reduced technical problems, but the inter-laboratory variation in Fpg-sensitive sites was not reduced. The ring-trials have not pinpointed specific assay steps as major determinants of the variation in DNA damage levels. It is likely that small differences in several steps cause inter-laboratory variation. Although this variation in reported DNA damage levels causes concern, ring-trials have also shown that the comet assay is a reliable tool in biomonitoring studies.