Caucasian and Oriental women have different incidence rates of breast cancer. Among the underlying risk factors for the development of breast cancer in the women of these two groups may be their different diets and patterns of estrogen metabolism and excretion. The absolute levels and relative ratios of 16 alpha-hydroxylated estrogens and 2-hydroxylated estrogens (catechol estrogens) in the body may have a role in the etiology of breast cancer, but studies so far have provided only conflicting results. Our goal was to study estrogen metabolism, in particular, the extent of 2-hydroxylation and 16 alpha-hydroxylation of estrogens in two groups of women, one Caucasian and one Oriental, with inherently different breast cancer risks. Dietary records were analyzed over 3-day periods in the mid-follicular phase, twice, at 6-month intervals for 13 premenopausal Oriental women, recent immigrant arrivals in Hawaii with presumed low risk of breast cancer, and for 12 premenopausal Finnish women with presumed higher risk. The urinary estrogen profile was measured by gas chromatography-mass spectrometry and plasma and fecal estrogens were assayed by chromatographic radioimmunoassays. Mean fat intake per 1000 kcal was 73% higher (P < .001) in the Finnish women, but the mean fiber intake and fecal weights were similar to those of the Oriental women. Compared with Oriental women, Finnish women had 46% higher plasma estradiol (P < .01) and 124% higher plasma estrone sulfate (P < .01); however, after adjustment for differences in age and body mass index, only the difference in estrone sulfate remained statistically significant (P < .05). Mean plasma levels of estrone and estradiol correlated with height after adjustment for body mass index (P < .05). Mean plasma levels of estrone and sex hormone-binding globulin were similar. The Finns had higher mean urinary estrone (193%), estradiol (166%), various catechol estrogens (130%-439%), and total estrogen excretion (123%) (all P < .001), but similar 16 alpha-hydroxylated estrogen excretion. As calculated, 16 alpha-hydroxylation of estrone was significantly increased (P < .01) in the Oriental women, but 2-hydroxylation, 4-hydroxylation, and 16 beta-hydroxylation of estrone were similar in both groups. The ratio of catechol estrogen to 16 alpha-hydroxylated estrogen was four to five times higher (P < .001) in the Finnish women. The Oriental women had two to three times higher fecal excretion of estrogens than the Finnish women (P < .01). Our results indicate that high catechol estrogen formation may be a greater risk factor for breast cancer than high 16 alpha-hydroxylation of estrogens. However, the main risk factor for the Finnish women, as opposed to the Oriental women, may be their higher estrogen levels that result from a higher fat diet, higher estrogen production related to their greater height, and lower fecal estrogen excretion.