Dietary and/or environmental factors appear to play a key role in the international variations that exist in breast cancer incidence. The genotoxicity of breast milk extracts is being examined as a possible indicator of in vivo exposure of mammary epithelial cells to DNA-damaging agents. Breast milk samples were obtained from the UK (n = 32), a high risk country, and from Hong Kong (n = 10), India (n = 20) and Singapore (n = 20), countries of lower breast cancer incidence. The abilities of breast milk extracts to induce DNA damage detected as single-strand breaks (SSBs) in the alkaline Comet assay and to induce micronuclei in MCL-5 cells and mutations in Salmonella typhimurium YG1019 were investigated. In the Comet assay 18 of 32 (56%) UK samples induced significant increases in DNA SSBs compared with 2 of 10 (20%), 5 of 20 (25%) and 8 of 20 (40%) of the samples from Hong Kong, India and Singapore, respectively. The proportion of positive samples was significantly higher in the UK group than in the combined low breast cancer incidence group and significantly higher than in the Indian group (P < 0.05, Fisher's exact test). In the micronucleus assay 9 of 32 (28%) UK samples showed significant activity compared with 0 of 10 (0%), 2 of 20 (10%) and 3 of 20 (15%) of the samples from Hong Kong, India and Singapore, respectively. Extracts of all the aforementioned milk samples were also tested for bacterial mutagenicity. Nine of 32 (28%) UK samples induced significant activity with a dose-response effect. Although activity was detected in samples from the other countries, comparable dose-response data could not be obtained because of a lack of material. This pilot study suggests that genotoxic components occur more frequently in UK breast milk than in milk from some other countries with a lower incidence of cancer. More work is required to confirm these initial findings and to examine their relevance to variations in breast cancer incidence.