THE CONCEPT of hypersplenism can theoretically be explained on the basis of a hormone-like function, overproduction of the hormone being the cause of disease states. Proof of this mechanism is lacking, particularly since animal experimentation on the pathologic physiology of the spleen has so far been unsatisfactory. Pregnancy complicated by hypersplenism, however, provides a clinical situation that can be interpreted as a controlled experiment, with comparison of the hematological condition of the mother at the time of delivery and that of the newborn infant. If the infant is born with the same disease as the mother but recovers spontaneously after separation from the maternal circulation, it would seem possible to postulate that the etiologic factor is a chemical that can cross the placental barrier. This fact was brought to our attention by a pregnant patient who at term experienced an acute purpuric crisis. A simultaneous splenectomy and cesarean section were