The availability of folate is implicated as a determinant of DNA methylation, a functionally important feature of DNA. Nevertheless, when this phenomenon has been examined in the rodent model, the effect has not always been observed. Several reasons have been postulated for the inconsistency between studies: the rodent is less dependent on folate as a methyl source than man; juvenile animals, which most studies use, are more resistant to folate depletion than old animals; methods to measure genomic DNA methylation might not be sensitive enough to detect differences. We therefore examined the relationship between folate and genomic DNA methylation in an elder rat model with a newly developed method that can measure genomic DNA methylation sensitively and precisely. Thirty-nine 1-year-old rats were divided into three groups and fed a diet containing 0, 4·5 or 18 μmol folate/kg (folate-deplete, -replete and -supplemented groups, respectively). Rats were killed at 8 and 20 weeks. At both time points, mean liver folate concentrations increased incrementally between the folate-deplete, -replete and -supplemented rats (P for trend <0·001) and by 20 weeks hepatic DNA methylation also increased incrementally between the folate-deplete, -replete and -supplemented rats (P for trend=0·025). At both time points folate-supplemented rats had significantly increased levels of DNA methylation compared with folate-deplete\ rats (P<0·05). There was a strong correlation between hepatic folate concentration and genomic DNA methylation in the liver (r 0·48, P=0·004). In the liver of this animal model, dietary folate over a wide range of intakes modulates genomic DNA methylation.