Background. The hypocholesterolemic effect of soy protein concentrates on normolipidemic subjects still remains unclear. Our objective is to assess the effect of soymilk supplementation, a whole soy product, with usual diet on serum lipids in normolipidemic subjects. Methods. We conducted a randomized controlled trial on 60 premenopausal normolipidemic Japanese women. After excluding 8 subjects whose initial serum concentration of total choleslerol or triacylglycerol was higher than 220 mg/dL (5.69 mmol/L) or 160 mg/dL (1.81 mmol/L), respectively, we encouraged the subjects in the soymilk-supplemented group (n = 27) to consume 400 mL (408 g) of commercial regular soymilk daily during two menstrual cycles. There were no significant differences in variables, including nutrient intake, between the soymilk-supplemented and control (n = 25) before the intervention. groups Results. After the trial, we observed a significant decrease of 10.9 mg/dL, or 5.3%, in serum concentration of total cholesterol in the soymilk-supplemented group. During the intervention, nutrient intake assessment showed significant increases in nutrient densities of vitamin E, polyunsaturated fatty acids, isoflavones, and P/S ratio and decreases in total energy and nutrient densities of vitamin C and green tea in this group. A statistically significant decrease in serum total cholesterol could still be observed even after excluding the estimated hypocholesterolemic effects of soymilk's polyunsaturated fatty acids. Conclusions. Our results suggest the hypocholesterolemic effect of soymilk, a traditional whole soy product, in Asian countries in normolipidemic subjects. (C) 2000 American Health Foundation and Academic Press.