Sir: The paper by Wilson et al. 1 on the relation between oestradiol levels and bacterial vaginosis (BV) provides further evidence for an hormonal influence in this condition. In 1996, or thereabouts, Phillip Hay and I made the suggestion that sex hormone changes might be involved in the aetiology of BV. 2 This was prompted not only by BV being noted, in observations over the course of a menstrual cycle or longer, to occur often in the first part of the cycle, as mentioned by Wilson et al., 1 but also by the results of experiments in mice. In such female animals, subcutaneous administration of oestradiol produced a BV-like picture in Gram-stained vaginal smears through multiplication of the endogenous vaginal bacteria. 3 In addition, this hormone was the key factor in promoting vaginal colonization by Mycoplasma hominis, ureaplasmas and Neisseria gonorrhoeae after their experimental inoculation. 4 -6 On the other hand, progesterone was the hormone that stimulated Chlamydia trachomatis and M. genitalium to grow in mice, oestradiol having no effect whatsoever in promoting growth of these microbes. 7,8 It is not difficult to see, however, that the patterns of hormonal effects are different in women from those in mice. In women high levels of oestradiol are associated with a normal vaginal flora and low levels with an abnormal flora, whereas mice given this hormone develop a BV-like picture and a susceptibility to some organisms, which otherwise is non-existent. The picture is BV-like in the sense that there is proliferation of endogenous murine vaginal bacteria, many adherent to epithelial cells, which, of course, are quite different from the bacteria found in the human vagina. Also different is the vaginal eco-system in women and mice. In women, the healthy vagina is dominated by acid-producing lacto bacilli causing the pH to be less than 4.5, rising to 7.0 or more in subjects with BV. In mice, lactobacilli are sparse and the pH is normally 7.4 (range 6.6-8.5) and not significantly different in mice given oestradiol, that is 7.36 (range 6.5-8.6), or progesterone, namely 7.2 (range 6.4-7.7). In addition, the nature of the cellular composition is somewhat different in women during the menstrual cycle than in mice during the oestrous cycle. In the former, a rising level of oestrogen is associated with proliferation of the vaginal stratified squamous epithelial cells and the surface cells become large and flattened with small or absent nuclei (cornification); in the second half of the cycle, with increasing progesterone, there is mucification of the squamous epithelium with large clear nucleated cells. Only at the end of the cycle, during menstruation, are erythrocytes and leucocytes seen. In the murine reproductive cycle, the oestrogenic phase is defined by squamous epithelial cells, some nucleated (early oestrous stage) or exclusively non-nucleated (oestrous stage) and a di-oestrous phase, stimulated by progesterone, characterized by an infiltration of polymorphonuclear leucocytes to the exclusion of epithelial cells...