Prenatal treatment of pregnancies at risk for congenital adrenal hyperplasia due to 21-hydroxylase deficiency was carried out in conjunction with chorionic villus sampling (CVS) in the first trimester for analysis of restriction fragment length polymorphisms. Dexamethasone administration to the pregnant woman was initiated at a mean gestational age of 7 weeks (range 4-10 weeks) before testing to determine whether the fetus was affected with 21-hydroxylase deficiency, and CVS was performed at a gestational age of 8-10 weeks. Two affected female fetuses were identified by molecular genetic techniques among this group. The duration of unnecessary prenatal dexamethasone treatment for unaffected or male fetuses was substantially reduced in the CVS group compared with a cohort of 8 prenatally treated pregnancies where amniocentesis was performed in the early second trimester. No major morbidities were observed in the treated pregnancies. Postnatal confirmation of CVS diagnosis was obtained in all cases where DNA from an affected sibling was available for comparative analysis with the DNA from chorionic villus tissue. The external genitalia of the affected females who were treated prenatally appeared normal. Based on these data we conclude that the benefit:risk ratio is favorable for prenatal administration of dexamethasone in pregnancies at risk for 21-hydroxylase deficiency. Treatment should be initiated during the first trimester in conjunction with diagnosis by CVS/molecular genetic techniques. Long-term postnatal surveillance is recommended for all offspring of dexamethasone-treated pregnancies.