The comet assay is now the method of choice for measuring most kinds of DNA damage in cells. However, due to the lack of a standardised protocol inter-laboratory comparisons are of limited value. The aim of this paper is to demonstrate how small changes in comet-assay variables may significantly affect the results. We examined the effect of varying agarose concentrations, alkaline unwinding time, electrophoresis time, voltage and current, by use of two cell types, viz, human peripheral blood lymphocytes and the lymphoblastoid cell line TK-6. All these variables have marked effects on assay performance and, therefore, on the determination of DNA damage. Here we identify factors of particular importance.