The comet assay (single-cell gel electrophoresis) is a sensitive and simple method for measuring DNA damage. An early modification of the assay involved the application of specific repair endonucleases to convert lesions to breaks; thus, for example, endonuclease III was used to measure oxidized pyrimidines. This concept has now been extended to produce an in vitro assay for DNA repair activity in a cell-free extract, for example from lymphocytes. The extract is incubated with substrate DNA containing specific base damage, and repair incision is detected as breaks in this DNA. We have recently been studying effects of phytochemicals in cultured cells, whether as antioxidants or as potential modulators of DNA repair. We realized that there is a need to check that observed effects that appear as an enhancement of repair (i.e. increased breaks in substrate DNA) are not simply due to a direct damaging effect of the phytochemical or to induction of non-specific nucleases. Here, we describe a rigorous approach to testing for this possibility, which we recommend to anyone carrying out similar experiments.