Methods are needed to assess exposure to genotoxins in humans and to improve understanding of dietary cancer prevention. The Comet assay was used to detect smoking-related exposures and dietary modulations in target tissues. Buccal scrapings, blood and faeces were collected from 38 healthy male volunteers (smokers and non-smokers) during a dietary intervention study with bread supplemented with prebiotics+/-antioxidants. GSTM1-genotype was determined with PCR. Buccal and peripheral lymphocytes were analysed for DNA damage using the Comet assay. Genotoxicity of faecal water (FW) was assayed in human colon HT29 clone 19A cells. 'Tail intensity' (TI) was used as a quantitative indicator of DNA damage in the Comet assay. Intervention with bread reduced DNA damage in lymphocytes of smokers (8.3+/-1.7% TI versus 10.2+/-4.1% TI, n=19), but not of non-smokers (8.6+/-2.8% TI versus 8.3+/-2.7% TI, n=15). Faecal water genotoxicity was reduced only in non-smokers (9.4+/-2.9% TI versus 18.9+/-13.1% TI, n=15) but not in smokers (15.5+/-10.7% TI versus 20.4+/-14.1% TI, n=13). The Comet assay was efficient in the detection of both smoking-related exposure (buccal cells) and efficacy of dietary intervention (faecal samples). Smokers and non-smokers profited differently from the intervention with prebiotic bread+/-antioxidants. Stratification of data by genotype enhanced specificity/sensitivity of the intervention effects and contributed important information on the role of susceptibility.