OBJECTIVE To determine the effect of maternal administration of dexamethasone on fetal heart rate and its variation. METHODS Retrospective analysis of computerised data derived from cases studied over three years. METHODS High risk pregnancy unit, John Radcliffe Hospital, Oxford. METHODS Twenty-eight pregnant women, at 27 to 32 weeks of gestation, to whom dexamethasone was given to accelerate pulmonary maturation in the expectation of preterm delivery. METHODS Dexamethasone (two doses of 12 mg intramuscularly, 12 h apart) was given on 51 occasions at weekly intervals (one to four occasions per patient). Complete data were available for cardiotocograph analysis from computerised measurement of fetal heart rate variables for two days before and four days after dexamethasone and, in 19 women, measurements of umbilical arterial flow velocity waveforms before and after dexamethasone. RESULTS In 10 pregnancies without fetal distress there was a highly significant (P < 0.01) transient rise in short term fetal heart rate variation after dexamethasone administration, from means (SE) 6.4 (0.28) to 9.8 (0.4) ms. In 18 pregnancies with subsequent delivery for fetal distress (abnormal fetal heart rate pattern) and high umbilical arterial resistance index [mean 0.93 (0.06 SE)], the rise in short term fetal heart rate variation was less (P < 0.01), from mean (SE) 5.4 (0.26) to 6.1 (0.48) ms. In a further case of discordant twin pregnancy, the larger twin continued to respond to dexamethasone administrations with a rise in fetal heart rate variation for five weeks; the smaller twin, with maintained tachycardia and reduced umbilical arterial end-diastolic flow velocity, failed to respond after the first two weeks. CONCLUSIONS The results show that maternal dexamethasone administration normally causes a rise in fetal heart rate variation for up to a day. This rise is reduced in pre-eclampsia or intrauterine growth retardation, associated with a reduction in umbilical flow, perhaps because of a consequential lower concentration of steroid in the fetus. The results contrast with those for betamethasone which has been reported to reduce fetal heart rate variation.