Volumes of feto-maternal transfusions (FMTs) in transabdominal chorionic villus sampling (TACVS) in the second trimester of pregnancy were calculated from the difference between maternal alpha-fetoprotein (AFP) concentrations before and 1 h after TACVS. In 50 pregnancies, there existed no correlation between FMT volume and the amount of villi collected or the number of TACVS attempts. The expected risk of fetal exsanguination due to very voluminous FMT could not be substantiated. In one case, immunization could have been the cause of hydrops fetalis, although only a volume of 0.15 ml could be calculated.