Although it is generally believed that a transplacental tap increases the risk of pregnancy loss associated with genetic amniocentesis, objective data are lacking. To quantitate the risk of transplacental amniocentesis, pregnancy outcomes were compared in 998 consecutive patients in whom amniocentesis was performed under direct real-time sonographic guidance. This technique allows visualization of the path of the needle during insertion and documents with certainty whether the placenta has been traversed. A transplacental approach was necessary in 347 patients (35%). The incidence of spontaneous abortion, antepartum stillbirth, prematurity (less than 37 weeks), and low birth weight (less than 2500 grams) was identical in patients undergoing both transplacental and nontransplacental amniocenteses. The similar rates of pregnancy loss in these two groups indicate that when midtrimester transplacental amniocentesis is ultrasonically guided, it is associated with little if any increased risk.