Exfoliated cells, isolated from breast milk samples donated by UK-resident women (n=15), were incubated, either immediately or after culture for 7 days, with one of a series of genotoxins, either in the presence or absence of the DNA-repair inhibitors, hydroxyurea (HU), and cytosine arabinoside (ara-C). The numbers of DNA single-strand breaks induced were then assessed as comet tail length (CTL) (μm) using the alkaline single cell–gel electrophoresis (‘Comet’) assay; cell viability was measured by trypan blue exclusion. The heterocyclic aromatic amines (HAAs) (2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) (0.4 mM), 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole (Trp-P-1) (1.67 mM), 3-amino-1-methyl-5H-pyrido[4,3-b]indole (Trp-P-2) (1.77 mM)), a polycyclic aromatic hydrocarbon (benzo[a]pyrene (B[a]P) (0.36 mM)), a nitro-polycyclic aromatic hydrocarbon (1-nitropyrene (1-NP) (1.84 mM)) and aromatic amines (o-toluidine (0.85 mM), p-chloroaniline (0.71 mM)) each induced statistically significant (P<0.0001, Mann–Whitney test) increases in median CTLs in breast milk cells from all the donors examined when incubated (30 min, 37°C) in the presence of HU/ara-C. In some cases, these compounds were also active in the absence of the repair inhibitors. There were marked variations in comet formation between donors and between genotoxins. Cell culture appeared to increase the epithelial cell proportion and cultured cells retained their ability to activate genotoxins. The results suggest that breast milk is a valuable source of human mammary cells for the study of the metabolic activation of possible carcinogens.