Two varieties of cassava, processed as Nigerian fermented flour (gari) or as Brazilian flour (farinha), were fed to two groups of eight infants and young children, each group receiving both forms of one variety, with preceding, intervening and following casein control diets. The flours provided 50% of diet energy, with casein added to make 8% energy as protein, vegetable oils to make 20% as fat and corn syrup solids and sugar to make 72% carbohydrate (CHO) energy. Fecal wet weight increased (P less than 0.05) from approximately 100 g/d (casein diets) to means (+/- SD) of 202 +/- 72, 171 +/- 58, 154 +/- 46 and 190 +/- 67 g/d; dry weights from means of 14.7-18.3 g/d to means of 22.9-24.4 g/d (P less than 0.05); fecal energy from means of 50-60 kcal/d to means of 89-94 kcal/d (P less than 0.01); fetal fat was generally not affected; and fecal CHO nearly tripled (P less than 0.01) from approximately 4 to 12 g/d. Apparent nitrogen absorptions and retentions from the cassava + casein diets were modestly lower than from casein diets. Rates of weight gain were very variable and not significantly different by diet; serum albumin levels were essentially unchanged. The results with these flours were indistinguishable from each other and from those previously found with freeze-dried cassava flour in otherwise identical diets. Variety and processing method had no effect on the digestibility of cassava starch and oligosaccharides and on the great resistance to digestion and the water-holding capacity of cassava fiber.