Pregnancy is a natural model of an optimal immune regulation in a graft-host relation. In order to satisfy contradictory interests of mother and foetus, a balance is established to protect the mother from infections and tumours and at the same time to prevent an immunological attack towards the foetus. In humans there is a well-established relationship between peripheral cytokine pattern and the outcome of pregnancy. It has been suggested that a Th2 dominant cytokine production favours the maintenance of normal pregnancy, whereas significantly increased Thl cytokine expression might represent the underlying phenomenon leading to pregnancy termination. The effect of progesterone on the immune system of pregnant women is at least in part receptor-mediated. Following recognition of fetal antigens, activated maternal g/d T cells express progesterone receptors, and upon progesterone binding produce a mediator; named PIBF. By signalling via a novel form of IL-4 receptor PIBF induces a TH2 dominant cytokine response, facilitating the production of IL-3, IL-4 and IL-10 by activated peripheral lymphocytes, and suppressing that of TH1 cytokines, such as IL-12 and IFN-g. Through altered cytokine production PIBF inhibits NK mediated killing in an indirect way. Neutralization of endogenous PIBF activity in pregnant mice by specific anti-PIBF antibody causes a significant reduction in the number of viable foetuses, and this is associated with an increased splenic NK activity, together with reduced IL-10 and increased IFNg production of the spleen cells. Ninety per cent of pregnancy loss is corrected by treatment of the pregnant animals with anti-NK antibodies. These data suggest that in mice PIBF contributes to the success of pregnancy and that the major part of its pregnancy-protective effect lies in keeping NK activity under restraint.