The passage of plasma proteins across the placental barrier in humans is known to be highly selective. Thus, free maternal insulin has been reported not to cross the normal maternofetal barrier, although insulin-binding antibodies have been detected in newborn infants whose diabetic mothers received insulin therapy. In this report we demonstrate, with the use of a human antiserum that permits distinction between human and animal insulins, that insulin in the cord blood of each of two neonates of insulin-treated diabetic mothers was, in part, animal insulin. The higher the antibody titer of the mother the greater was the total insulin in the cord plasma and the greater was the fraction that was animal insulin. In case 1 cord plasma insulin was 0.7 unit/liter, of which 10% was animal insulin; in case 2 cord plasma insulin was 3.5 units/liter, of which 25% was animal insulin. The demonstration that antigen restricted from transplacental passage can be transferred while complexed to antibody raises the question whether such fetal exposure would induce partial or total immunologic unresponsiveness subsequently if the fetus were rechallenged with the same antigen.