Pre-eclampsia is a frequent, unpredictable syndrome which is dangerous for both mother and foetus. The concept of placental ischemia has gained wide acceptance among the numerous theories put forward to explain the illness. The setting up of preeclampsia seems to be scheduled in two steps: (1) an absolute or relative placental ischemia due to vascular diseases or hypertrophic placenta, or most often secondary to implantation defect, particularly anomaly with the invasive trophoblast; (2) a diffuse endothelial disease. The connection between these two steps is incompletely disclosed. The authors demonstrate that the maternal immune system which is strongly stressed during all the stages of normal gestation is implicated in pre-eclampsia. Its role is probably not univocal. Foeto-trophoblastic antigens could be poorly recognised. This defect of recognition could lead to the abnormalities of trophoblastic invasion observed in pre-eclampsia. Pre-eclampsia does not seem to be accompanied by an immunological rejection of the foetus. Some genetically predisposed patients do not have a sufficiently competent immune system to neutralise one or more of the toxic products released by the ischemic placenta. Certain types of pre-eclampsia could be auto-immune, with the auto-antibodies directed against certain types of phospholipids or trophoblastic constituents. A disequilibrium between oxidation and anti-oxidation mechanisms involving neutrophils could lead to aggression of the endothelium which is observed in pre-eclampsia. Pre-eclampsia could represent a form of immuno-dystrophy, with the excessive production of adverse cytokines locally, directed against the trophoblast. Without directly implicating the immune system as the trigger of pre-eclampsia, it seems that its role is unclear. In some cases it develops protective mechanisms which, when overwhelmed or inadequate, allows pre-eclampsia to occur. In other cases it can form part of the cascade of aggressions leading to the abnormalities encountered. The integration of these abnormalities in the pathophysiological models, could help improve the classification of pre-eclampsia. This attempt will lead to a more adapted preventive and therapeutic management of pre-eclampsia.