Folate bioavailability of beef liver, lima beans, peas, spinach, mushrooms, collards, orange juice and wheat germ was estimated with a protocol of folate depletion-repletion using growth and liver, serum and erythrocyte folate of weanling male rats. Diets with 125, 250 and 375 micrograms folic acid/kg were standards. Individual foods were incorporated into a folate-free amino acid-based diet alone (250 micrograms folate/kg diet from food) or mixed with folic acid (125 micrograms folate from food + 125 micrograms folic acid) to evaluate folate bioavailability and effects of food matrix. Beef liver and orange juice folates were as available as folic acid, whereas those of wheat germ were less bioavailable. Folates of peas and spinach were also less available than folic acid using liver and serum folate concentrations and total liver folate as response criteria, but they were not lower when based on growth and erythrocyte folate concentrations. Lima bean, mushroom and collard folates were as available as folic acid using four of five response criteria. Folate bioavailability of all foods generally exceeded 70%. All response criteria gave approximately equivalent results, indicating that growth and tissue folate levels are appropriate criteria. No food matrix effects were observed for any food except lima beans. Foods rich in polyglutamyl folates were less bioavailable than those of foods rich in short-chain folates.