In a study designed to estimate the requirement for dietary folate in nonpregnant women, 17 women (21-27 y) consumed 200, 300, or 400 micrograms/d of total folate for 70 d which was provided by low folate conventional foods (30 micrograms) plus supplemental folic acid. Group means for initial serum and erythrocyte folate and plasma homocysteine concentrations were not significantly different. Serum and erythrocyte folate decreased relative to the initial value in the 200 micrograms/d group (43.4 +/- 12.1%, 13.6 +/- 16.6%, respectively; mean +/- SD), in contrast to an increase in the 400 micrograms/d group (16.8 +/- 52.0%, 10.2 +/- 18.5%, respectively). The final serum folate in the 200 and 300 micrograms/d groups (6.4 +/- 0.8 nmol/L, 7.3 +/- 1.1 nmol/L, respectively) was significantly lower than that of the 400 micrograms/d group (14.3 +/- 2.0 nmol/L), with evidence in the 200 micrograms/d and 300 micrograms/d groups of low ( < 6.8 nmol/L) serum folate concentrations. Differences in final erythrocyte folate did not reach statistical significance, although low values ( < 362 nmol/L) were frequent in subjects with 200 micrograms/d intake. In the 200 micrograms/d group, plasma homocysteine was negatively correlated with serum and erythrocyte folate, and final mean plasma homocysteine (12.6 +/- 1.7 mumol/L) was significantly higher than that of the 300 or 400 micrograms/d groups. Elevated plasma homocysteine levels ( > 16 mumol/L) were observed in the 200 micrograms/d group only.(ABSTRACT TRUNCATED AT 250 WORDS)