The authors start with an anatomical description which makes it possible to distinguish from one another septate, bicornuate unicornuate and pseudo-cornuate uteri. The diagnosis is often suggested by the clinical features and is made by using complementary examinations such as ultrasound, hysterography, hysteroscopy and laparoscopy. Although it is difficult to be sure of the incidence of the condition it is thought to be around 1%. In the 10 years from 1978 to 1987 the authors have studied 296 pregnancies occurring in 109 patients who were not operated on and 34 pregnancies occurring in 21 patients who had plastic operations on their uteri. They show that the condition has an effect at all stages of pregnancy and delivery and in the first category of patients (abortion in 38%, extra-uterine pregnancy in 8%, prematurity, pre-eclampsia and malpresentation in 28%, with a high rate for caesarean section in 37%). The prognosis was quite different in patients who had been operated on. The published literature confirms these findings. The authors discuss the different ways of carrying out surgical treatment (simple section or excision of the septum, or Bret-Palmer's plastic operation on the uterus with treatment of the associated vaginal abnormalities), plus the medical treatment (with rest and the use of vasodilatators and drugs for anti-platelet aggregation) to be used during the pregnancy. They conclude that it is necessary to make a diagnosis of what type of malformation is present, to decide which surgical procedures to carry out as this does improve the prognosis greatly.