In their Letter to the Editor, Lillford et al. have questioned the analyses and conclusions of the results presented in our recently published paper (1). In this study, mice were exposed to tetrachloroethylene at the doses of 0 (vehicle control), 1000 and 2000 mg/kg/day. From the results obtained, we concluded that tetrachloroethylene is weakly genotoxic in the liver of mice. This conclusion was not shared by Lillford et al. as it, in their view, was heavily dependent on statistical analysis and did not take biological relevance of the results into account. We do not agree with this opinion and, therefore, the criteria and arguments we have taken into account in the evaluation of the results are further elaborated below. In toxicity testing, it is generally held that the finding of a dose-related effect of a test substance provides a very strong argument for concluding that the observed effect is biologically relevant. In the study we have published, it was noticed that the group mean values for tail intensity in the liver increased with the dose, suggesting a biologically relevant effect of tetrachloroethylene, i.e. induction of DNA damage. This positive trend was then analysed with the Jonckheere–Terpstra test and found to be statistically significant. Thereby, the statistical analysis has been used as an aid in evaluating the results, and the outcome of the analysis showed that there was a low probability for the observed positive trend to occur by chance only. We disagree with the comment by Lillford et al. stating that our conclusion was heavily dependent on statistical analysis and that statistical analysis was the sole criterion for a positive effect. The biological relevance of the effect, demonstrated by the dose-related increase in tail intensity in the liver, was considered first and has been taken into account for determining the positive response. Lillford et al. argue that groups of mice treated with tetrachloroethylene exhibited tail moments and tail intensities that were similar to those of the concurrent vehicle control group. Given the flexible meaning of the word similar and that the DNA-damaging effect of tetrachloroethylene was weak, we understand the basis for such a statement. However, 8 of the 12 tetrachloroethylene-treated animals had higher tail intensity values in the liver than the animal with the highest tail intensity value in the concurrent vehicle control group and, furthermore, the dose-related increase in tail intensity established in the liver implies that the mean tail intensity values for the three experimental groups were indeed different. We note that Lillford et al. comment that the maximum group mean increase was only 1.43-fold (tail intensity in the liver, 2000 mg/kg group) above the vehicle control group and that the increase in tail moment in this group was only 1.36-fold. In our opinion, the crucial result of the study was that a weak DNA-damaging effect of tetrachloroethylene was established. Whether the magnitude of this weak effect was 1.43-fold or 1.36-fold is of less importance. Furthermore, the reference to tail moment is of limited value since according to international expert panel recommendations for conducting the alkaline in vivo comet assay (2), the use of tail intensity (i.e. per cent DNA in tail) is recommended, as indicated by the following statement: