The immune system probably plays a role in the onset and development of endometriosis. A general picture can be proposed. In some women refluxing endometrial cells are not destroyed, either because the patient is genetically programmed not to respond to endometrial antigens, or because the reflux is so abundant that the scavenging capacity of the peritoneal immune cells is overloaded. Refluxing cells could be protected due to an abnormal adherence to the mesothelium which exceptionally expresses certain adhesive molecules. Undestroyed, these endometrial cells would cause an inflammation with activation of macrophages. Not only does the peritoneum protect these endometrial cells, but it also produces abnormal quantities of chemotactic and angiogenic cytokines (interleukin-8). Macrophages facilitate development via growth factors such as transforming growth factor P. Immunosuppressive factors block the cytotoxic activity of natural killer (NK) cells. Activated macrophages present antigens of endometrial cells to T cells which will co-operate with B cells to synthesize autoantibodies. Synthesized antibodies protect the ectopic endometrium and could worsen the dysfunction of local NK cells. A vicious circle is set up involving all the partners of the immune system. It is as yet impossible to pinpoint the triggering mechanism. The primary defect could be localized on the endometrium, macrophages already activated by an extrinsic factor (infection, spermatozoa, chemical substances), the uterus or the tubo-uterine junction. The two pathophysiological theories put forward to explain endometriosis are linked by a defective immune system. Indeed, once the vicious circle is set up, growth and angiogenic factors could induce metaplasia of the already irritated mesothelium.