Most investigators agree that in the early-onset form of neonatal group-B streptococcal disease the infecting micro-organism is acquired from the maternal genital tract (Harper, 1971; Baker and Barrett, 1973; Franciosi, Knostman and Zimmerman, 1973; Anthony and Okada, 1977; Ferrieri, Cleary and Seeds, 1977; Anthony, Okada and Hobel, 1978). Prematurity, prolonged rupture of the fetal membranes, and other obstetric complications are more frequent in colonised infants who develop disease than in those who do not (Baker and Barrett, 1973; Franciosi et al., 1973). None of these factors influences the development of late-onset group-B streptococcal disease, in which nosocomial transmission may be of more importance (Steere et al., 1975; Aber et al., 1976). The factors that influence colonisation of neonates are not well understood. In a few studies in which the sex, race, length of gestation and birth weight of asymptomatic colonised and non-colonised infants were compared, no differences were found (Baker and Barrett), 1973; Aber et al., 1976; Baker, 1977). Some infants who do not yield group-B streptococci when swabbed immediately after birth have acquired them by the time of discharge from the hospital (Steere et al., 1975; Aber et al., 1976; Ferrieri et al., 1977). Whether they are at the same risk as infants colonised at birth of developing early or late-onset disease is not known. If high rates of neonatal colonisation favour an increased incidence of disease (Regan et al., 1979), and because 28 to 72% (Aber et al., 1976; Baker, 1977; Regan et al., 1979) of infants born to mothers with positive genital cultures acquire group-B streptococci, it is important to identify factors that might favour the transmission of the organisms from mother to infant. We have attempted to do this in a prospective study of a population of mothers and their newborn infants.