Following the administration of diazepam to mothers in labour the clinical effects, changes in thermogenesis, and metabolism of the drug in the newborn were observed under carefully controlled conditions. A total maternal dose of 30 mg or less in the 15 hours before delivery had little effect on the infants' state. Larger doses, however, were accompanied by low Apgar scores at birth, apnoeic spells, hypotonia, reluctance to feed, and an impaired metabolic response to a cold stress. Measurement of plasma levels of diazepam and its active metabolite showed that both products were detectable in significant concentrations in some infants for up to eight days. We conclude that greater care should be taken in the use of this otherwise effective drug for the treatment of pre-eclampsia.