Although there are a number of studies and individual case reports concerning the use of benzodiazepines in human pregnancy, the data concerning teratogenicity and effects on postnatal development and behaviour are inconsistent. There is evidence from studies in the 1970s that first trimester exposure to benzodiazepines in utero has resulted in the birth of some infants with facial clefts, cardiac malformations, and other multiple malformations, but no syndrome of defects. Diazepam and chlordiazepoxide are amongst the drugs most frequently implicated in the earlier studies. However, data from later studies provide no clear evidence of significant increase in either the overall incidence of malformations or of any particular type of defect. Many of the women included in these studies has psychiatric illnesses, epilepsy, or diabetes all of which have an intrinsic risk in pregnancy, and some were on multidrug therapy. Medical-obstetric histories and family history of malformations were not always presented in the publications, which makes assessment of risk associated with benzodiazepine use per se difficult. Nevertheless, in most of the studies involving first trimester use of benzodiazepines, the majority of infants were normal at birth and had normal postnatal development. Late third trimester use and exposure during labour seems to be associated with much greater risks to the fetus/neonate. Some, but by no means all infants exposed at this time, exhibit either the floppy infant syndrome, or marked neonatal withdrawal symptoms. Symptoms vary from mild sedation, hypotonia, and reluctance to suck, to apnoeic spells, cyanosis, and impaired metabolic responses to cold stress. These symptoms have been reported to persist for periods from hours to months after birth. This correlates well with the pharmacokinetic and placental transfer of the benzodiazepines and their disposition in the neonate. However, there has been no significant increase in the incidence of neonatal jaundice and kernicterus in term infants. The prolonged use of benzodiazepines throughout pregnancy raised the concern that there may be altered transmitter synthesis and function, leading to neurobehavioural problems in the children. In approximately 550 children who were followed up for various times up to four years of age, there is no increase in either the malformation rate or adverse effects on neurobehavioural development and IQ. Although some of the data indicate that a small number of children were slower to develop during the first year or so, they did exhibit catch up growth and most had developed normally by four years of age. Where developmental deficits persisted, it was not possible to prove a cause-effect relationship with benzodiazepine exposure. These children were often from families where there was maternal illness requiring prolonged drug therapy or where there were social problems. It is important to consider poor environmental and social factors when assessing the possible prenatal influence of the benzodiazepines on the postnatal health and development of the child. There is evidence that clonazepam, clorazepate, diazepam, lorazepam, midazolam, nitrazepam, and oxazepam are excreted into breast milk. The published data indicate that the levels detected in breast milk are low; therefore, the suckling infant is unlikely to ingest significant amounts of the drug in this way. Problems may arise if the infant is premature or has been exposed to high concentrations of drug either during pregnancy or at delivery.