Peritoneal fluid and the intraovarian milieu are a specific microenvironment. Peritoneal fluid originates mainly as an ovarian exudation product caused by increased vascular permeability, with cyclic variation in volume and steroid hormones which are always higher than in plasma. It contains large amounts of macrophages and their secretion products, and has a large exchange area with plasma through the peritoneum, which is highly permeable for small molecules. Diffusion becomes virtually zero for molecules with a molecular weight of >100000 Da. In women with the luteinized unruptured follicle (LUF) syndrome, concentrations of oestrogens and progesterone are much lower in the luteal phase. Endometriosis is associated with sterile low-grade inflammation, increased concentrations of activated macrophages and many of their secretions, such as cytokines, growth factors and angiogenic factors. Concentrations of CA-125 and of glycodelins are also increased, secreted locally by the endometrial cells. Natural killer (NK) cell function declines, possibly mediated by glycodelins or local intercellular adhesion molecule (ICAM) -1 shedding. The ovary is also a specific microenvironment, with steroid hormone concentrations 1000-fold higher in follicles than in plasma. Endometrial and superficially implanted cells are influenced by peritoneal fluid concentrations so that local environment, rather than inherent cellular differences could explain differences between superficial endometriosis and eutopic endometrium. Differences between superficial implants and endometriotic disease, deep infiltrating or cystic ovarian endometriosis, may thus arise via different endocrine environments. Superficial endometrial implants are regulated by peritoneal fluid factors, whereas deep endometriosis and cystic ovarian endometriosis are influenced by blood or ovarian factors. The endometriotic disease theory considers superficial endometriotic implants and their remodelling as a physiological process in most women, and concentrates on the causes of severe endometriosis such as differences in the eutopic endometrium from women with and without endometriosis (which may indicate hereditary differences), the invasiveness of some endometriotic cells in vitro, focal 'shielding' of endometriotic foci by adhesions, and inhibition of NK activity by ICAM-1 and glycodelins. Endometriotic disease is thus seen as a benign tumour. The type of cellular lesion, hereditary and immunological environments and local hormone concentrations in the ovary and in peritoneal fluid, will decide expression as cystic ovarian endometriosis, deep endometriosis or adenomyosis externa, and whether the latter is associated with adhesions.