The normal diet of six healthy volunteers was supplemented by 21 g of dietary fiber from two different soybean seed fiber preparations, either a nonpurified and never-dried soya pulp A (39% dietary fiber) or a purified soya fiber B (79%), for 3 wk each. Mean daily fecal wet weight was increased by 19 and 38% in the fiber periods A and B as compared to a 2-wk control period (p less than 0.05). Stool frequency and transit time remained unchanged. Fecal fiber increased by 52% only during period B. Fecal calcium, magnesium, and iron were increased (p less than 0.05), mainly after higher intakes during both fiber periods. Excretion of neutral steroids remained unchanged, but their concentration was lowered (p less than 0.05). A 21% increase of fecal bile acids by fiber A was specifically due to deoxycholic acid (+32%, p less than 0.01). Oral glucose tolerance was slightly improved after period B. Neither fiber changed serum triglycerides, but fiber B increased low-density lipoprotein-cholesterol by 19% (p less than 0.01) and low-density lipoprotein-phospholipids by 16% (p less than 0.05). The ratio high-density/low-density lipoprotein-cholesterol, however, did not change significantly. Thus, dietary fibers from soybean do not seem to contribute to the hypocholesterolemic effect of soya. The results of this study also demonstrate that two apparently similar dietary fibers, coming from a single source, can exert distinctly different metabolic effects.