Investigators have found an association of low birth weight with the isolation of genital mycoplasmas from the maternal cervix and urine. In other studies, Ureaplasma urealyticum and Chlamydia trachomatis have been implicated in nongonococcal urethritis, currently the predominant sexually transmitted disease. The lower genitourinary tract is the probable route of access of these organisms to the endometrium and the fetal membranes. The present study was designed to investigate the presence of urea-plasmas, mycoplasmas, chlamydiae, fungi, aerobic and anaerobic bacteria, and cytomegalovirus in fetal membranes and to determine their association, if any, with perinatal morbidity and mortality. The placentas of three groups of subjects were cultured: fetuses or neonates which died in the perinatal period, newborns admitted to the intensive care unit, and normal controls. The perinatal-death group included 75 fetuses which were stillborn after at least 20 weeks of gestation and 69 infants who died after birth. The intensive care group consisted of all infants (452 in number) admitted to the intensive care unit, whose birth weights were 1500 gm or less. One live-bom infant was selected as a control for each member of the perinatal-death group. Ureaplasma and mycoplasma were isolated most frequently from the placentas of intensive care infants. Also, isolation rates among the placentas of stillborn and dead infants were approximately double those among the controls. Twenty-five per cent of such placentas harbored either or both of the microorganisms. The variation in the isolation rate of any mycoplasma among the four groups (counting stillborn and dead infants as separate groups) was significant (P < 0.0007). Isolation of group B streptococci did not differ significantly among the four groups (P = 0.18). Cytomegalovirus was isolated once from the placenta of an intensive care infant. Neither C. trachomatis nor Mycoplasma pneumoniae was isolated from the cnoriontc surface. Although aerobic and anaerobic bacteria and fungi were isolated, they were present infrequently among the four groups. In subsequent analyses, the stillbirth and infant-death groups were combined as one perinatal-death group. The association between placental cultures positive for U. urealyticum and birth weight is shown in Table 1. Among all subjects, recovery of ureaplasma from the placenta showed a strong and stable inverse relationship with birth weight. Similar results were obtained when the data were analyzed according to gestational age. Among all infants, relative to those born at 36 or more weeks of gestation, the odds ratios for U. urealyticum were: 3.6 for 20–27 weeks, 2.8 for 28–31 weeks, and 1.7 for 32–35 weeks. An appropriate relationship between birth weight and gestational age was found in both ureaplasma-positive and ureaplasma-negative subjects. Upon histological examination, inflammation of the membranes was seen in 60 per cent of the 58 placentas from which ureaplasma was isolated. Twenty-nine per cent of the membranes with cultures negative for ureaplasma showed inflammation. This difference in the occurrence of chorioamnionitis was significant (P = 0.0001). A significant excess was observed for inflammation of the umbilical cord also. No association was found between U. urealyticum and inflammation of the chorionic villi. Maternal diabetes and preeclampsia were negatively related to colonization with U. urealyticum. The following maternal characteristics were positively related to such colonization: race other than white or black, age less than 25 years, single status, student status, primigravidrty, spontaneous rupture of membranes, membrane rupture 48 or more hours before delivery, and status as a cigarette smoker at time of delivery. Little or no association was seen between colonization with U. urealyticum and the following characteristics: prior pelvic inflammatory or venereal disease, consumption of alcohol or coffee, prior spontaneous or induced abortion, method of payment, or