OBJECTIVE: To establish identifiable prenatal factors in fetal heart block which might predict death in utero, the need for intervention, or the probability of pacemaker requirement. SETTING: Tertiary referral unit for fetal echocardiography. SUBJECTS: 36 fetuses with congenital complete heart block and structurally normal hearts identified between 1980 and 1993. METHODS: Maternal anti-Ro antibody status was documented. Prenatal variables examined included absolute heart (ventricular) rate, change in rate, and development of hydrops fetalis. Postnatally, heart rate, need for pacing, and the indications for pacing were detailed. RESULTS: Of the total of 36 patients, there are 24 survivors; 11 are paced. Of those fetuses which died, two were electively aborted for severe hydrops, seven died in utero, two were immediate postnatal deaths, and one was an unrelated infant death. The trend was for the heart rate to decrease during fetal life and postnatally. Fetuses with deteriorating cardiac function did not always show the lowest heart rates. Bradycardia of less than 55 beats/min in early pregnancy or rapid decrease in heart rate prenatally were poor prognostic signs. Hydrops was also associated with bad outcome, 10 out of the 12 hydropic fetuses dying (83%). Of 10 fetuses presenting with a heart rate above 60/min, nine survived of whom three required pacing. Of seven presenting with heart rates of 50/min or less, only three survived and two of these required pacing. Of the two fetuses with negative maternal anti-Ro antibody status one died in utero and one required heart transplantation after pacemaker insertion. CONCLUSIONS: Isolated complete heart block identified in fetal life does not always have a good prognosis. An individual heart rate does not accurately predict the outcome in utero or the need for postnatal pacing. Regular, careful monitoring during pregnancy is required in order to optimise care and timing of any interventions.