Objective: The hypothesis was tested that the rate of postmenopausal bone loss is inversely associated with long-term urinary excretion of phyto-oestrogens, as a marker of habitual dietary intake. Design: Secondary analysis of a 10-year follow-up study (1979-1989) among postmenopausal women in the Netherlands. Subjects: From the original population of 154 women, 32 women were selected with an annual rate of radial bone loss of ≤ 0.5% over the first 5 years of the study and 35 women with a rate of ≥ 2.5% per year. Methods: The isoflavonoids genistein, daidzein and equol, and the lignan enterolactone were determined by gas chromatography - mass spectrometry in aggregate samples from annually collected urine samples. Cortical bone density of the radius had previously been measured annually by single-photon absorptiometry. Results: Excretion of isoflavonoids did not differ between both groups, although in multivariate analysis equol excretion was weakly positively associated with rate of bone loss in the 5 years after the menopause. Enterolactone excretion was significantly higher in the group with high rate of bone loss. This positive association remained in multivariate linear regression analysis after adjustment for age, years since menopause, body mass index and intake of calcium, vegetable protein and dietary fibre. Conclusions: Enterolactone excretion is likely to be an indicator of consumption of grains and legumes; it is not clear whether the observed positive association with rate of bone loss is a causal one. Our results do not support a preventive effect of low, unsupplemented dietary intake of phyto-oestrogens on postmenopausal cortical bone loss. However, no conclusions can be drawn about effects of higher doses of phyto-oestrogens.