Abnormalities (dysmorphism, mental retardation etc) were reported (The Lancet 1987; i: 108-9) in children of mothers using high doses of benzodiazepines regularly during pregnancy in the area of Sweden with the highest use of benzodiazepines (Gothenburg). This signal was investigated in Sweden by use of available drug data bases, and in the USA by use of computerized health care data in Medicaid, including drugs and diagnoses. Neither the Swedish survey, nor the preliminary analysis of the pediatric profiles in the US-Medicaid population supported the occurrence of drug-associated abnormalities as described in The Lancet. Also, no significant findings were seen in a survey in Czechoslovakia on behavioural changes in children, whose mothers used diazepam during the second part of their pregnancy. Although available pharmacoepidemiology data do not confirm the hypothesis, this review points to existing geographic differences in use of benzodiazepines during pregnancy (frequency of chronic use and dosage level) and their potential clinical consequences. (Fig. 1, Ref. 25).