his project includes three multicenter trials of the safety and accuracy of prenatal diagnosis procedures, chorionic villus sampling and amniocentesis: 1) a nonrandomized comparison of Transcervical (TC) CVS and amniocentesis procedures in 8,000 obstetrical patients at 7 clinics which found an estimated excess risk of pregnancy loss after TC CVS compared with amniocentesis of 0.8% (NEJM, 1989); 2) a randomized Transabdominal (TA) vs. Transcervical (TC) CVS study of 4000 patients at 8 clinics which revealed low overall pregnancy loss rates and indicated that when practitioners are experienced, the two methods are equally safe (NEJM, 1992); 3) a study of 3,000 Danish patients randomized between amniocentesis, TA and TC CVS which found risk of fetal loss was similar after TA CVS and amniocentesis procedures, but was considerably greater following TC CVS (Lancet, 1992). NICHD Cooperative Agreement, 1984-1992.