The binding of myasthenia gravis antibody to acetylcholine receptor (AcChoR) as measured in vitro by radioimmunoassay with 125-labe ed a-bungarotoxin (a-BuTx), can be blocked by amniotic fluid, maternal serum, and umbilical cord serum. This inhibitory effect is due to a-fetoprotein present in high concentrations in amniotic fluid and serum, as shown by: (i) selective removal of several components from amniotic fluid and serum; (ih) selective addition of different components present in amniotic fluid and serum, including a-fetoprotein, to the radioimmunoassay; (iii) correlation between the inhibitory effect of both amniotic fluid and serum and between the amounts of a-fetoprotein they contain; (iv) blocking of the a-fetoprotein effect by pretreatment of the amniotic fluid with anti-a-fetoprotein anti'y. The inhibitory effect of a-fetoprotein in vitro suggests a similar effect in vivo in pregnant women with myasthenia gravis. This effect may explain in part the variability in the development of neonatal myasthenia gravis in the babies, due to transplacental transfer of maternal anti-AcChoR antibody, only after delivery and only in the minority of the cases. It also may explain the appearance of remissions in females with myasthenia gravis during the second and third trimesters of pregnancy. Similar phenomena observed during pregnancy in other autoimmune and immunopathogenic diseases also might be attributed to activity of a-fetoprotein.In studies from our laboratory (1, 2), we demonstrated that the in vitro binding of antibodies to acetylcholine receptor (AcChoR) from patients with myasthenia gravis to AcChoR antigen can be inhibited by: (i) human amniotic fluid obtained from healthy women during the second trimester; (Hi) serum obtained from healthy women during the second and third trimesters; and (iii) umbilical cord serum obtained after delivery from both healthy women and myasthenics. We suggested that a similar inhibitory effect in vivo on the binding of specific antibodies to AcChoR at the neuromuscular junction may occur during pregnancy in patients with myasthenia gravis. Because myasthenia gravis is an autoimmune disease caused by the immunopathologic effect of specific autoantibodies directed against AcChoR at the neuromuscular junction (3), such inhibitory effects may help to explain two phenomena observed in myasthenia gravis during pregnancy and after delivery: The occurrence of partial or complete remissions during the second and (especially) the third trimester, as well as exacerbations during the early postpartum period (4-8); and the appearance of transitory neonatal myasthenia gravis (NMG) in the newborn babies, due to placental transfer of maternal antibodies (9-11), only after birth and only in a minority of the cases (12, 13). This paper presents experimental data showing that the immunosuppressant a-fetoprotein, but not other components derived from both human amniotic fluid and human umbilical cord, is the factor inhibiting the binding of antibodies to AcChoR antigen present in the...